Evaluating controls on coupled hydrologic and vegetation dynamics ...

Evaluating controls on coupled hydrologic and vegetation dynamicsin a humid continental climate watershed using a subsurface-landsurface processes model

Chaopeng Shen,1 Jie Niu,2 and Mantha S. Phanikumar2

Received 20 November 2012; revised 23 February 2013; accepted 8 March 2013; published 28 May 2013.

[1] Understanding key controls on hydrologic dynamics is important for effectivelyallocating resources for data collection, reducing model dimensionality, and makingmodeling decisions. This work seeks to elucidate the physical factors responsible for theobserved hydrologic patterns of a watershed in Michigan using an integrated hydrologicmodel, Process-based Adaptive Watershed Simulator and Community Land Model(PAWSþCLM). The model is tested using observed data for streamflows, soil temperature,groundwater table depths, and satellite-based observations of evapotranspiration (ET) andleaf area index (LAI). Numerical experiments are carried out to lump the effects of keycontrols, including land use types, nitrogen levels, groundwater redistribution, and soiltexture, into different process indices. Using analysis of variance (ANOVA), wequantitatively determine the strengths of these controls on ET, net primary production(NPP), and other important variables. Groundwater flow is found to be the major control onrunoff and infiltration, with soil texture ranking next, while vegetation type and nitrogenlevels are found to dominate NPP, top soil temperature, and transpiration. Soil texture andgroundwater are found to have comparable influence on soil moisture, which is inagreement with analysis of field data in the literature. All controls are found to colimit ET,which serves as the nexus for ecosystem-hydrology interactions. From the simulationresults, we find that nitrogen significantly controls transpiration, through which it influencesother hydrologic fluxes. While there is room for improving descriptions of the nitrogencycle in the current version of CLM, these novel results call for an understanding of theinterplay between hydrology and biogeochemistry. Additional analysis shows that therelative strengths of the controls examined in this work are fairly robust with respect tochanges in parameters and spatial resolution.

Citation: Shen, C., J. Niu, and M. S. Phanikumar (2013), Evaluating controls on coupled hydrologic and vegetation dynamics in a humidcontinental climate watershed using a subsurface-land surface processes model, Water Resour. Res., 49, 2552–2572, doi:10.1002/wrcr.20189.

1. Introduction

[2] Recently, there has been a surge of interest in identi-fying, understanding, and utilizing hydrologic and ecosys-tem patterns and controls in hydrologic predictions [Hwanget al., 2012; McDonnell et al., 2007; Thompson et al.,2011; Vivoni, 2012; Wagener et al., 2010]. Understandingthe physical processes that control the variability of hydro-logic variables is crucial for predictions and management[Joshi and Mohanty, 2010; Pan and Wang, 2009; Wilsonet al., 2004]. The recognition of key patterns and controlscan help reduce the dimensionality of the prediction prob-lem [Sivapalan et al., 2011], facilitate predictions in unga-

uged basins [Sivapalan, 2003], and promote an in-depthunderstanding of the underlying principles. Topography-induced groundwater flow leads to convergence of subsur-face moisture, which crafts spatial patterns in hydrologic,geomorphologic, and vegetation processes. In some sys-tems, groundwater flow plays a vital role in ecosystemfunctioning. However, despite intensive recent studies[e.g., Ivanov et al., 2008; Maxwell and Kollet, 2008a,2008b; Miguez-Macho and Fan, 2012], the relative impor-tance of groundwater flow as compared to other controlssuch as soil texture, land use, and vegetation characteristicsremains unclear.

[3] The coupling between hydrology and biogeochemis-try has recently received growing attention [Chorover etal., 2011; Lohse et al., 2009; Manzoni and Porporato,2011; Riveros-Iregui et al., 2012]. Strong controls andfeedbacks have been identified between water, carbon, andnitrogen biogeochemistry in various environments [Burtand Pinay, 2005; D’Odorico et al., 2003; Lohse et al.,2009; Schimel et al., 1997]. The strong relationshipbetween evapotranspiration (ET) and biological nitrogenfixation (BNF) [Cleveland et al., 1999] has long been rec-ognized, so has the correlation between nitrogen

1Department of Civil and Environmental Engineering, PennsylvaniaState University, University Park, Pennsylvania, USA.

2Department of Civil and Environmental Engineering, Michigan StateUniversity, East Lansing, Michigan, USA.

Corresponding author: C. Shen, Department of Civil and EnvironmentalEngineering, Pennsylvania State University, University Park, PA 16802,USA. ([email protected])

©2013. American Geophysical Union. All Rights Reserved.0043-1397/13/10.1002/wrcr.20189

2552

WATER RESOURCES RESEARCH, VOL. 49, 2552–2572, doi:10.1002/wrcr.20189, 2013

availability, successional stage, soil moisture, and topogra-phy [Robertson et al., 1988]. In ecosystems where vegeta-tion growth is limited by nutrient availability [Galloway etal., 2004; Wilson et al., 1999], higher inputs of nitrogenare expected to lead to a relatively more active ecosystemwith higher productivity, thus altering ET and the hydro-logic cycle. Presently, human activities are creating reac-tive N at unprecedented rates [Galloway et al., 2004],leading to increased nitrogen deposition [Dentener et al.,2006; Galloway et al., 2008]. The implications of thisexcess nitrogen and its interplay with the carbon and watercycles is worth detailed, mechanistic and integrated investi-gation. The relative importance of N is difficult to evaluatefrom field surveys alone due to the various time scales ofthe complex N biogeochemical reactions, its interactionswith the carbon cycle, and its covariance with climaticinput [e.g., see Burke et al., 1997]. On one hand, inclusionof nitrogen dynamics in global land surface simulations isrelatively recent (see Bonan and Levis [2010], Reich et al.[2006], and Sokolov et al. [2008] for the effects of the C/Ninteractions), and the current hydrologic descriptions inthese climate scale models are very crude, oversimplifyingthe subsurface and channel dynamics. On the other hand,with the exception of very few (TOPOG-IRM [Vertessy etal., 1996] and RHESSys [Band et al., 2001; Mackay andBand, 1997; Tague and Band, 2004]), ‘‘traditional’’ water-shed-scale models often do not consider any carbon/nitro-gen dynamics. Understanding the relative strengths of thewater-carbon-nitrogen coupling helps identify unrecog-nized yet important linkages.

[4] Besides synthesizing and mining observed data forcorrelations, one may also study the influences of separateprocesses using the simulations of physically based hydro-logic models (PBHMs), which are ‘‘derived deductivelyfrom fundamental physical laws’’ [Beven, 2002]. AlthoughPBHMs may never match the full complexity of reality,they have the potential to reveal patterns and pinpointcause-effect relations. To reconcile differences and similar-ities in formulations and to aid scientific understanding ofthe complex processes involved, new models, approaches,and model intercomparison exercises are all expected toplay an important role. Recent advances in PBHMs havefocused on the linkages of terrestrial fluxes, ecohydrology,and the integral role played by surface and subsurfacewater dynamics. For an incomplete list, see the papers byFatichi et al. [2012], Ferguson and Maxwell [2010],Goderniaux et al. [2009], Ivanov et al. [2008], Kollet andMaxwell [2008], Mackay and Band [1997], Miguez-Machoet al. [2007], and Niu et al. [2013]. A novel hydrologicmodel, Process-based Adaptive Watershed Simulator(PAWS), was recently introduced by Shen [2009] and Shenand Phanikumar [2010, hereinafter SP10]. PAWS was cre-ated with intermediate complexity, Geographical Informa-tion System (GIS) data interface and parameterizationfunctionalities, attempting to strike a balance betweenphysics and computational tractability at large scales. Themodel efficiently solves the governing equations for majorhydrologic processes using some of the best available algo-rithms. By reducing the dimensionality of the fully three-dimensional (3-D) subsurface problem using a quasi-3-Dsaturated groundwater domain and one-dimensional Rich-ards’ equation to approximate soil columns in every grid

cell, the model significantly reduces the computationaldemand with little loss of physics. The computational effi-ciency of the PAWS model allows for long-term, large-scalesimulations and makes parameter estimation and uncertaintyanalysis feasible. The PAWS model was tested extensively(SP10) with laboratory ‘‘recharge’’ experiments, analytical solu-tions, idealized test cases, and numerical solutions from modelsthat solve the full 3-D Richards equation. The model achievedgood performance in simulating hydrology in a medium-sizedwatershed in the U.S. midwest. Descriptions of land surfaceprocesses, including energy balance and vegetation growthcycles, were nonetheless simple in the original PAWS model.

[5] The National Center for Atmospheric Research(NCAR) community land model (CLM), a process-basedland surface model, was developed from grassroots collab-oration among climate scientists [Collins et al., 2006;Dickinson et al., 2006; Lawrence et al., 2011; Niu andYang, 2007; Oleson et al., 2010; Sakaguchi and Zeng,2009; Zeng et al., 1998, 2002]. The model now encom-passes a comprehensive suite of land surface processesincluding surface heat/momentum/vapor transfer, surfaceradiation balance, snow/soil heat transfer and freeze-thawphase changes, photosynthesis and plant growth, as well ascarbon and nitrogen fluxes, mostly using process-baseddescriptions. However, the flow modules in CLM, e.g., sur-face runoff and channel flow, subsurface flow, and charac-terization of geologic and soil properties, all tend to beoverly simplified as compared to other components. CLMtreats each computational element as independent columns,neglecting the lateral fluxes and hence their spatial interac-tions. As discussed in Anyah et al. [2008], Maxwell andKollet [2008a], and Shen and Phanikumar [2010], the sub-surface flow is an integral component of the hydrologiccycle that significantly influences local water fluxes and theclimate through feedbacks. To further enhance the simula-tion capabilities of the PAWS model, we recently coupledthe flow processes in PAWS to the land surface processesin CLM. This suite of modules brings together surfaceflow, 3-D subsurface flow, carbon/nitrogen dynamics, andland surface processes to offer a higher degree of modelrealism.

[6] In this paper, we employ the PAWSþCLM model tostudy the water-energy-nutrient controls on hydrology in ahumid continental climate watershed in the Great Lakesregion of North America. Although progress has recentlybeen made in understanding the hydrology of this regionwith a focus on the impacts of land cover/land use changes[Mao and Cherkauer, 2009] and projected climate changeon hydrology [Mishra et al., 2010], the applications ofPBHMs have the potential to offer additional insights intofundamental processes. Our objectives in this paper are (a)to quantify the relative importance of physical processesincluding land cover, soil texture, land use and ground-water flow on water, energy, and carbon fluxes using a pro-cess-based hydrologic model and (b) to examine thecoupling strengths and mutual influences between hydrol-ogy and nitrogen biogeochemistry. The paper is organizedas follows. After a brief description of the PAWSþCLMmodel, we present the results of extensive model testingusing field and satellite-based observations. Using the well-tested model, we show the rich spatiotemporal patterns ofthe hydrologic and vegetation dynamics in the basin. Then,

SHEN ET AL.: EVALUATING CONTROLS ON HYDROLOGIC AND VEGETATION DYNAMICS

2553

we carry out numerical experiments to analyze the relativeimportance of the controlling processes on various fluxesand states and how they together produce the patterns.

2. Model Description

[7] Since the mathematical and algorithmic details of thePAWS model have already been presented elsewhere(SP10), we will only provide a brief outline of the modelhere followed by short descriptions of the coupling withCLM. PAWS uses 3-D structured grids with the top layerrepresenting the ground surface and the overland flowdomain. Several key processes take place exclusively onthe ground surface (e.g., overland flow, vegetation intercep-tion, depression storage) while others (e.g., ET) extendover the entire vertical depth of the soil column. To avoidconfusion, we define our terminology as follows: the term‘‘cell’’ is used to refer to the unit of the 2-D horizontal dis-cretization of the ground surface (this corresponds to the‘‘grid cell’’ concept in CLM 4.0). We use the word ‘‘col-umn’’ to denote the totality of the soil matrix and porespaces underneath a cell (under each cell, we only have onecolumn, with a single set of moisture/thermodynamicstates, as opposed to the possibility of multiple columns inCLM); the word ‘‘layer’’ refers to the vertical discretiza-tion of the soil column.

2.1. Flow Modules

[8] The flow domain is divided into surface overlandflow, channel flow, unsaturated soil water flow, and satu-rated groundwater. Processes in these domains are linkedby the surface-unsaturated-saturated coupling methoddescribed in SP10. The ground surface is partitioned intothe flow domain and the ponding domain. The pondingdomain is updated together with the soil water states. Run-off rate from the ponding domain to the flow domain isdetermined by the Manning’s equation. Only water inexcess of the interception depth can become runoff. Underflooding conditions, water in the flow domain may alsobackfill into the ponding domain. For the overland flow do-main, we solve the 2-D diffusive wave equation using anefficient and stable Runge-Kutta finite volume (RKFV)scheme. A similar method is used for the channel network,which exchanges flow with the overland and the ground-water flow domains. The exchange between groundwaterand channel flow is calculated based on the leakance con-cept [Gunduz and Aral, 2005], after the diffusive waveequation is solved. Wetlands are an important land covertype that modifies hydrologic responses. A new lowland-storage module is developed as part of this work and isdescribed in Appendix A (Figure 1 illustrates the concep-tual model used by the module).

[9] The subsurface is discretized into a series of 1-D soilcolumns connected to saturated groundwater layers at thebottom. The saturated-unsaturated soil water flow is gov-erned by the Richards equation, which is solved using amodified Picard iteration approach [Celia et al., 1990; vanDam and Feddes, 2000]. The vadose zone module handlesinfiltration into the soil under normal conditions; however,under heavy rainfall conditions, we employ a modifiedform of the generalized Green and Ampt method [Jia,1998]. We bring the dynamics of regional groundwater

flow into the soil columns by writing a separate mass bal-ance equation for the last grid cell, whose thicknesschanges as the water table fluctuates. The bottom of the lastcell extends to the top of the bedrock. Lateral flowexchange calculated by the groundwater flow equation isfed into the Richards equation, and the recharge fluxobtained from the solution of the Richards equation is inturn passed to the groundwater flow equation as a sourceterm. As described in SP10, this coupling method producesphysically consistent (in that the soil moisture profile isconsistent with the groundwater head) and stable solutionsand compares well with solutions based on the fully 3-Dvariably saturated Richards equation models for differentconditions (e.g., conditions corresponding to infiltrationexcess, saturation excess, and return flow).

2.2. (PAWS1CLM): Coupled Subsurface and LandSurface Processes

[10] The processes included in CLM have been detailedin Oleson et al. [2010]. For the sake of completeness, weprovide a brief overview of CLM processes incorporated inthe coupled model in this section and then discuss linkages.PAWS is coupled to the latest release of CLM (version4.0), with a prognostic crop model [Lawrence et al., 2011].The soil hydrology and river routing routines in CLM arereplaced by the corresponding procedures in the PAWSmodel. The coupling details are described in section 2.2.2.2.2.1. Introduction to CLM Processes

[11] For canopy radiative fluxes, the two-stream approxi-mation of Dickinson [1983] and Sellers [1985] is used tokeep track of incoming, transmitted, reflected and absorbeddirect or indirect radiation, over multiple solar wavebands.The plants are parameterized by detailed ecophysiologicaldescriptors to capture the optical variability among plantspecies, especially trees. Canopy scaling/integration of sun-light penetration is based on the specific leaf area (SLA,the ratio of leaf area to leaf mass) concept [Thornton andZimmermann, 2007], which is based on the assumption that

Figure 1. Illustration of the lowland-storage module.This compartment is conceptualized as a storage compart-ment of the overland flow domain.


2554

SLA increases linearly as a function of overlying leaf areaindex (LAI), and the two-big-leaf canopy model, whichtreats the canopy as a sun-lit leaf and a sun-shaded leaf[Dai et al., 2004]. CLM employs the resistance concept,rather than the widely used Penman-Monteith (PM)approach to calculate momentum/heat/vapor transfer.Although the PM equation assumes a wet-bulb approxima-tion for vegetation surface temperature, the resistanceapproach explicitly solves for the leaf temperature to sat-isfy the coupled latent and sensible heat transfer equations.Aerodynamic resistances are calculated based on theMonin-Obukov similarity theory [Kundu and Cohen, 2010;Zeng et al., 1998]. Canopy resistances are solved simulta-neously with photosynthesis using the Farquhar model forC3 plants [Farquhar et al., 1980] or the model of Collatz[Collatz et al., 1992] for C4 plants. As classified in Arora[2002], this is a biochemical approach that provides a pro-cess-based description of CO2 assimilation and may bemore suitable for the assessment of future climate scenar-ios. The water stress function couples hydrology with eco-system dynamics and is parameterized as described insection 2.2.2.

[12] Soil and snow temperatures are updated by solvingthe heat conduction equation. The SNICAR (SNow, ICe,and Aerosol Radiative model) module [Toon et al., 1989]included in CLM 4.0 keeps track of the mass balance ofaerosols and impurities in the snowpack. Snow agingimpacts effective grain size and consequently snow albedo.Soil freezing is calculated by a freezing-point depressionformulation of Niu and Yang [2006]. This formulationallows supercooled water and ice to coexist in a wide rangeof below-freezing temperatures. The soil hydraulic conduc-tivities are scaled down by an impermeability factordepending on the fraction of ice in the soil water content.

[13] The carbon and nitrogen cycle on the land surfacesare fully incorporated with the biogeochemical vegetationdynamics and phenology module of Biome-BGC [Thorntonand Rosenbloom, 2005; White et al., 1997]. The photosyn-thesized carbon is sent into a central carbon pool, which isfirst spent on maintenance respiration by live compartments(leaf, fine root, live coarse root, and live stem if applica-ble). Surplus carbon is then allocated to growth and pheno-logical carbon pools (leaf, dead steam, root, and storage/transfer pools) depending on the allometric relationshipsprescribed for different plant species. This process isstrongly controlled by available nitrogen. Growth respira-tion is consumed when allocation occurs. Offsetting, gapmortality, and fragmentation of coarse wood debris sendcarbon and nitrogen into the litter pools. Carbon and nitro-gen (C/N) then trickle down a cascade of litter/soil organicmatter (SOM) pools with varying turnover rates. The dis-played structure of vegetation (LAI, tree height) is prognos-tically updated based on the carbon stored in differentpools. Different plant species are parameterized with differ-ent phenological traits (e.g., evergreen, stress-deciduous,seasonal deciduous, etc). Nitrogen transformations arebook kept in plant pools, three litter pools, four SOM pools,and a soil mineral pool. Simulated nitrogen-specificdynamics include deposition, fixation, leaching, and uptake.

[14] It is worth mentioning that although CN (Carbonand Nitrogen) module in CLM is a complex model withrespect to the number of pools as compared to many other

models [Manzoni and Porporato, 2009], there are still largeuncertainties with the biogeochemical cycling in CLM.The CN module is a compartment model and does not trackchemical species (except for the distinction between cellu-lose and lignin pools), microbial community, age/residencetime, and vertical distribution of C/N, all of which can playimportant roles [e.g., see Berg and Laskowski, 2005; Fanget al., 2005; Maggi and Porporato, 2007; Maggi et al.,2008; Wutzler and Reichstein, 2008; Xu-Ri and Prentice,2008]. The allometric coefficients, base turnover rates(base rates are then adjusted by temperature and waterfunctions), nitrogen deposition/fixation rates, and denitrifi-cation rates all face uncertainties. There are also uncertain-ties associated with the appropriate spatiotemporal scales atwhich these rates should be applied [Manzoni and Porpor-ato, 2009]. Data sets to validate the internal dynamics ofCN cycling models are generally lacking.2.2.2. Coupling CLM With PAWS

[15] The coupling between PAWS and CLM is done in amass-conservative, theoretically consistent, and tightlyintegrated fashion. At initialization, vertical and horizontaldiscretization information in PAWS, which depends on to-pography, soil, and geology, is ported to CLM so that bothhave the same definitions of computational units. At eachtime step, climate forcing is passed into CLM to computeinterception, throughflow, surface radiation processes, sur-face energy balance, photosynthesis, soil temperature, andsnow processes. Then, the ET demand (both transpirationand soil evaporation) as a function of depths is passed intoPAWS as a source term for the unsaturated zone. Alsopassed along are the soil temperature, ice content, canopystorage, ground precipitation, dew, and snowmelt amounts.The soil hydrology module in CLM is replaced with itsPAWS counterpart, which then solves the Richards equa-tion together with ET, groundwater flow, and runoff basedon the coupling scheme described in SP10. Overland flowand streamflow are also calculated in PAWS. The resultingsoil moisture states are then converted into the soil waterstate variables in CLM and supplied back into CLM forecosystem updates and calculations in the next time step.

[16] In this paper, we will focus on the results obtainedby running the (PAWSþCLM) model in the ‘‘offline’’mode, in which the climate forcings are based on observa-tions from conventional land-based stations. PAWS uses astructured grid; therefore, the CLM discretization is config-ured such that there is a one-to-one mapping between aPAWS cell and a CLM grid cell, and there is only one col-umn in each CLM grid cell. At the same time, subcell het-erogeneity on the plant functional types (PFT) level issupported as PAWS also adopts a similar approach. InSP10, this subcell heterogeneity of land use/land covertypes was called representative plant types (RPT). To con-form to the CLM literature, we adopt the term PFT in thispaper. In the coupled model, PAWS provides the front-endprocessing utilities that reclassify raw land use data setsinto CLM PFTs.

[17] Some changes to the CLM processes are necessaryto make the two models consistent in theory. Since the soilwater flow processes are primarily computed in PAWS, thecombined model uses the van Genuchten formulation forsoil water retention relationships as in the original PAWSmodel. Field capacity, saturated water content, and wilting


2555

points are set by the PAWS model. The soils characteriza-tions are derived from the SSURGO database (Soil SurveyGeographic database from the Natural Resources Conser-vation Service [Soil Survey Staff, 2010]), which is a highresolution, comprehensive data set. The soil resistances toevaporation are reformulated according to Neitsch et al.[2005]. The root water uptake efficiency (or root resistancein CLM) adopted by PAWS, that is, the Lai and Katul[2000] formulation, is used. This formulation contains atunable parameter � (Table 1) that modulates the root effi-ciency-water stress function. When coupled to atmosphericmodels, the climate input data is normally forced at 30 mabove canopy. With land-based climatic stations, however,the observation height is between 1 and 2 m. Therefore, theaerodynamic resistances to vapor/heat/momentum transferare formulated using a neutral stability assumption. Whenwind flow velocity is low (<1 m/s), free convection isassumed, and the formulation in Ivanov [2006] and Kondoand Ishida [1997] is employed.

[18] Climate-forcing data are passed into CLM fromPAWS for the calculation of energy and water fluxes. Theresulting ET fluxes are passed back to PAWS and con-verted into source terms for the Richards equation. We alsoconsider phase change in the subsurface. The freezing-point depression formula in Niu and Yang [2006] is utilizedin the vadose zone model in PAWS to reduce the hydraulicconductivities in freezing soils. The urban and dynamicland cover modules in CLM 4.0 are not currently used.

[19] Some minor adjustments include the following: tomatch with literature values in Hessl et al. [2004], the fineroot carbon/nitrogen ratio of broadleaf deciduous forest(DBF) was changed from 45 to 60 and the leaf carbon/nitro-gen ratio of temperate needleleaf evergreen forest waschanged from 25 to 45; both C3 grass and C3 crop havebeen changed to seasonal deciduous to avoid a problem withthe CLM parameterization of stress-deciduous PFTs that leadto spurious growth in brief periods of warmth in winter.

3. Results

[20] In this section, we apply the PAWS model to amedium-sized watershed located in the U.S. LaurentianGreat Lakes region. The model is first tested using observa-tions and then utilized to elucidate the essential hydrologicdynamics of this watershed. Although extensive validationsare provided to demonstrate that the new model candescribe key processes accurately, the focus of this paper isin evaluating the relative strengths of controls on coupled

hydrologic and vegetation dynamics. After describing thebackground hydrology, we use the model to study the rela-tive importance of several controlling processes for surfacefluxes: land use and vegetation type, soil texture, topogra-phy-induced groundwater flow, and nitrogen levels.

3.1. Site Description and Input Data

[21] The Clinton River (CR) watershed (1837 km2) is situ-ated in the mid-southern part of the lower peninsula of Michi-gan, as shown in Figure 2, and drains into Lake St. Clair.Located in the humid continental climate (hot summer)region, the basin experiences large seasonal temperature var-iations with very warm summer (average July temperaturesabove 22�C) and cold winter (average January temperaturesaround �3�C). Precipitation is evenly distributed in the yearwith an annual average of approximately 900 mm, which isslightly higher compared to nearby regions due to the lakeeffect. Daily or subdaily weather data are obtained from theNational Climatic Data Center (NCDC) [2010]. The loca-tions of the weather stations are shown in Figure 2a.

[22] This watershed is selected because of its diversity inland use and topography. The topography of the watershedmainly consists of rugged hills on the highlands of the westand flat, low-lying plains in the east, joined by a steep east-facing slope. The southwest corner forms semienclosedsubbasins with numerous small lakes. The 30 m resolutionnational elevation data set (NED) is preprocessed to gener-ate average cell elevation and lowland-storage bottom ele-vation in the computational grid. As shown in Figure 2b,the watershed is urbanized in the southern portion withvarying extent of modification. The northern half is occu-pied by forest in the western quarter and agriculture in theeastern quarter. The 30 m resolution IFMAP (IntegratedForest Monitoring, Assessment, and Prescription) 2001land use land cover data [Michigan Department of NaturalResources, MNDR, 2010], which was classified from Land-sat Thematic MapperTM imagery is used to provide landuse information. Three dominant land use types (PFTs) aremodeled in each horizontal cell. The soil color data areextracted from the global data set [Global Soil Data Task,2000].

[23] The surficial aquifer systems of the watershed aremainly composed of Pleistocene glacial drift that overliesthe bedrock. Due to its proximity to Lake St. Clair, thewatershed is covered by less transmissive lacustrine fine-grained sediment in the nearshore plain, while the highlandportion of the watershed is characterized by relatively morestratified and permeable outwash. The glacial drift istreated as the unconfined aquifer. Spatial fields of the con-ductivities (lateral) of glacial drift were obtained by inter-polating well records from the WELLOGIC database[Groundwater Inventory and Mapping Project (GWIM),2006; Oztan, 2011; Simard, 2007] using kriging after noisewas removed. Also interpolated from the WELLOGICdatabase are the aquifer thicknesses and static water tableheights to provide discretization information and compari-son data for the model. The bedrock is mainly coldwatershale, which is composed of shale and some limestone.This is a confining unit layer with very low permeabilitythat produces little water yield. Lithology-based estimatesof bedrock conductivities (lateral) are rasterized and inputinto the model. Lake St. Clair has been found to influence

Table 1. Model Calibration Parameters

Symbol (Unit) Parameter Meaning

K (m/day) Groundwater hydraulic conductivityKS (m/day) Soil saturated hydraulic conductivityN van Genuchten parameterA (1/m) van Genuchten parameter� Parameter in Lai and Katul [2000] root efficiency

function�ice Scale-dependent freezing fraction parameter as � in

equation (11) in Niu and Yang [2006]Kr (m/day) River leakancesl (m) Length of flow path for runoff contribution to

overland flow domain


2556

groundwater flow in the basin. To better approximate real-ity, we enforce a constant head boundary condition at theboundary face nearest to Lake St. Clair. The average lateralconductivity of the unconfined aquifer in the basin is 13 m/day.

[24] The model was spun-up by repeatedly loopingthrough the climatic forcing data between 2001 and 2009over a span of 4000 years, as in Thornton and Rosenbloom[2005]. To reduce the computational time required formodel spin-up, several typical sites in the watershed are

chosen. The spin-up was done in a ‘‘column mode,’’ inwhich the phreatic water table depth is kept constant. It hasbeen verified that the ecosystem has approximately reachedequilibrium. The resulting carbon and nitrogen states andother ecophysiological parameters are stored to be read inat run time.

[25] A 880 m � 880 m horizontal grid is used for thecomputations on the landmass. Twenty vertical layers areused for the unsaturated soil zone. The vertical discretiza-tion of the soil layer at a location depends on the local

Figure 2. (a) Map of the Clinton River watershed showing its location, elevation, weather stations,USGS gages (labeled by eight digit numbers), and the hillslope transect (A-B). (b) Land use map of thewatershed.


2557

thickness of the vadose zone and uses an adaptive approachthat assigns finer grid near the surface and coarser cellsnear the bottom. Fourteen major rivers are simulated in themodel, all discretized with spatial grid sizes of approxi-mately 1500 m (A flexible exchange scheme in PAWSallows for flexible discretization of channel and land). Theland surface and subsurface processes run on a maximumtime step of 1 h. Overland flow and channel flow use a maxi-mum of 10 min time steps. All components can adaptivelychange step sizes as necessary. On average, a day of serial-ized simulation takes 2.5 s on an Intel 2.3 GHz i-7 CPU.

3.2. Model Testing Against Observations

[26] The majority of the parameters in PAWS are spa-tially distributed (e.g., hydraulic conductivity or soil prop-erties), and their spatial heterogeneity is honored in oursimulations. Modest adjustments are done to some parame-ters, such as the van Genuchten soil parameters, by apply-ing a global multiplier to improve model performance.Parameters adjusted are listed in Table 1. USGS (U.S. Geo-logical Survey) gaging station 04165500 at Mt. Clemens,which represents basin outflow, is selected to show thecomparison between model simulations and observationsfor the watershed. A modified Nash-Sutcliffe coefficient,MNASH is used as a model performance metric :

MNASH ¼ 1� NASHþ RNASH

2;

where NASH is the standard Nash-Sutcliffe coefficient, andRNASH is the Nash-Sutcliffe computed based on square-

root transformed data (SP10). The rationale for usingMNASH is that the NASH coefficient tends to give toomuch importance to the peaks (runoff) at the expense ofimportant subsurface processes (e.g., baseflow contributionto streams). Use of the MNASH metric is expected toaddress this deficiency of the original NASH coefficient.Figure 3a shows the comparison between simulated andobserved daily hydrographs at the outflow USGS gage, CRat Moravian Drive at Mt. Clemens, MI. Figure 3b shows aclose-up of a portion of the hydrograph. In Figure 3c, wealso show the hydrograph comparison at an inner gage, CRat Sterling Heights, which is not involved in calibration.The decent daily NASH values (0.61 and 0.65) at bothbasin outlet and inner gages are satisfactory. Qualitatively,the baseflow is captured very well. The model grosslyunderestimates the largest peak in May 2004, while later inJune 2004 there is a separate, large simulated peak thatdoes not correspond to any observed peaks. We suspectthat some error with precipitation data has led to this tem-poral mismatch, which causes heavy penalty in NASH.

[27] Simulated depths to water table (averaged from2007 to 2009) are compared with the measured values fromthe WELLOGIC database in Figure 4 (the model is startedin 2001). As can be seen from the spatial fields and thehigh R2 value (0.66), an overall good agreement is notedbetween simulated and observed water table depths.

[28] Figure 5 shows the simulated soil temperature at alocal measurement site in Romeo, Michigan [Enviro-weather, 2010]. The main deviation is in the winter of2008, when the model predicts as low as �5�C while thetemperature sensor reported around 0. This is attributed toa combination of factors, including potentially imperfectsoil freezing scheme, local conditions, and measurementerrors. Overall, the model was able to reproduce the majordynamic cycle of soil temperature fluctuations. The formu-lations for soil freeze and thaw should continue to improveas this process heavily influences surface water and energyfluxes [Sinha and Cherkauer, 2008]. However, we need tokeep in mind that the simulated variable represents an aver-age over the model grid cell (880 m � 880 m) while the

Figure 3. Simulated daily streamflow compared withUSGS measurements at basin outflow gage (a, b) and aninner gage (c).

Figure 4. Observed versus simulated depth to watertable. The observed data are obtained by kriging data in theWELLOGIC database.


2558

observed data are measured at a point, which can beexpected to deviate from the average.

[29] Figure 6a shows the comparison of simulated LAIwith the level 4, 8-day Terra MODIS LAI product(MOD15A2) [Knyazikhin et al., 1999; NASA Land Proc-esses Distributed Active Archive Center (LP DAAC),2012]. In Figure 6a, the urban areas (where MOD15A2gives NaN values) are blanked out and shown in white foreasier visual comparison with the MODIS observations.Figure 6b shows the time-series comparison of simulatedand observed (MODIS) LAI at two locations with relativelyuniform land use type of DBF and soybean, respectively.We experience some problems when comparing withMODIS LAI: (a) MODIS products always show no LAI atall in winter while the model does predict LAI for ever-green forest (ENF), which ranges from 5.5 to 6.5 (but canbe covered by snow) and (b) peak values of LAI for cornpublished in the literature can reach 5–6 [e.g., Howell etal., 1997; Suyker and Verma, 2009], which matches withthe simulated peak corn LAI, whereas the MOD15A2 LAIfor the agricultural regions in this basin never goes above 3.Both problems lead to simulated LAI being higher thanMODIS observed LAI. However, the literature values seemto suggest this may be a problem of the MODIS rather thanthe model. We notice that the LAI of the DBF and soybeanare very well simulated, especially the onset and peak periods(the MODIS product does show some temporal variation ofpeak LAI for the DBF, which is more likely due to cloudinterferences and observational errors). The simulated offsetof DBF is much more abrupt than the observed, which is dueto the linear offset function used in the tree phenology inCLM, suggesting that a more smooth offset formulationshould be considered in future model enhancements.

[30] Figure 7a shows the simulated ET versus theMODIS-based ET estimates (MOD16A2 and MOD16A3[Mu et al., 2011, 2007]). The large-scale spatial pattern ispreserved, including relatively lower ET in the eastern agri-cultural region and higher ET in the central-western for-ested areas. However, some fine-scale patterns predicted bythe model were not observed in the MOD16A2 data. Forexample, the model simulates a high ET pocket at the cor-ner of the southwestern boundary, whereas the MODIS val-ues there are very low. The values cannot be this low,

given that the region is filled with inland lakes. The simu-lated agricultural ET tends to be higher than MODIS prod-ucts. Considering the uncertainties and errors with theMODIS product, we argue that this level of match is quiteencouraging. Figure 7b shows the temporal comparison ofbasin-average ET between simulated and MOD15A2. Thesimulated ET tends to be slightly higher than the MODISproduct for most of the months. In general, the temporaldynamics agree very well.

3.3. General Background Hydrology of the Watershed

[31] The primary purpose of this section is to use themodel to elucidate the essential hydrologic dynamics in the

Figure 5. Observed versus simulated daily soil tempera-ture at Romeo, Michigan.

Figure 6. Comparison between MODIS product(MOD15A2, 8 day) and the PAWSþCLM simulated LAI.(a) Spatial map comparison for June 2005 and December2005. (b) Time series comparison for two locations withrelatively uniform land cover type of deciduous broadleafforest (DBF) and soybean, respectively.


2559

watershed, which serves as the context for the analysis ofrelative importance in section 3.4. As we will see, the spa-tiotemporal patterns observed in this watershed are gov-erned by different processes with variable strengths ofcontrols. These patterns naturally drive us to ask the ques-tions about the relative strengths of controls.

[32] Insight into the hydrologic cycle of this watershedcan be gained from Figures 8 and 9, which show, respec-tively, mean monthly basin-average fluxes, temporal trendof the total soil moisture content, TSMC (including waterand ice), and snow water equivalent, SWE. Qgc, Ex, SatE,and InfE are the four major mechanisms of streamflow gen-eration, namely, groundwater contribution to streamflow,lowland exfiltration, saturation excess, and infiltrationexcess. Their sum can be seen as approximate, but notequal, to the streamflow, because of interception andreinfiltration.

[33] Figure 8 vividly describes a system that is energycontrolled most of the year and becomes moisture limitedonly in the hottest summer months. Precipitation increasesfrom March until it reaches its peak in May. Streamflownormally reaches its peak in March. The first few precipita-tion events can be very pronounced in terms of streamflowgeneration. In April and May, although precipitationincreases more than 80%, ET jumps even more rapidly andovercomes the added precipitation. Correspondingly, theTSMC of the basin shifts into the ‘‘net-loss’’ phase (Figure

9), and the soil moisture pool starts to be tapped to compen-sate for the increasing ET demand. There is still amplemoisture in the subsurface so that streamflow does not dropnoticeably. From June to August, we observe a rapiddecline of soil moisture content in the basin. Soil sel-dom reaches saturation. ET shifts from energy-limited tomoisture-limited conditions. Streamflow consequentlydrops to its annual minimum in August. From Septem-ber to the end of the year, ET wanes quickly whenTSMC steps into the ‘‘net-gain’’ region. ET almost van-ishes in winter and the soil continues to store wateruntil the next cycle starts.

[34] The energy-driven dynamics is also clear from thestreamflow generation. The overland flow contribution tostreamflow is commonly divided into saturation excess(SatE) and infiltration excess (InfE) [Dingman, 2008],which together create the peaks in the hydrograph. Wedefine SatE as runoff that occurs when the entire soil

Figure 7. Comparison between MODIS products andPAWSþCLM simulated evapotranspiration (ET). (a) Spa-tial map comparison of annual ET (mm) in 2005. (b) Time-series comparison of basin-average monthly ET (mm).

Figure 8. Monthly mean streamflow generating fluxesand ET (mm) from 8 years of simulation. The black linewith diamond symbol is precipitation.

Figure 9. Basin average state variables showing interan-nual variations of (a) soil water content (change from initialstate) and (b) snow water equivalent. The year 2003 is adrought year.


2560

column underneath the surface area is saturated. Figure 8demonstrates that streamflow generation is tightly relatedto the seasonal energy cycle. Saturation excess evolves inphase with TSMC in Figure 9 and is the dominant processin winter and spring. The dominance of SatE again impliesthat the system is energy limited in cold seasons. Infiltra-tion excess, on the other hand, is less important than SatEin cold seasons, which is explained by the high conductiv-ity of glacial drift soils in the region. The relative impor-tance of InfE rises in summer, when streamflow diminishes

and river stages are lowest. In August, storm events barelystimulate much streamflow, as TSMC is very low and thebasin is far from saturation, and the small flashy peaks thatdo occur are mostly composed of infiltration excess, asshown in Figure 8b. InfE is primarily generated from theurbanized regions of the southern basin and, when SatEdiminishes, its percentage in the runoff rises.

[35] Figure 10 illustrates the spatial patterns of long-termaverage annual energy/water states and fluxes. The watervariables seem dominated by topography, whereas the

Figure 10. Annual average spatial distributions of (a) latent heat flux (W/m2); (b) sensible heat flux(W/m2); (c) top 10 cm soil temperature (�C); (d) top 10 cm soil moisture (e) infiltration excess (mm/yr)(runoff generated before soil column is saturated) ; (f) saturation excess (mm/yr) (runoff generated attime of saturation); (g) recharge to groundwater (mm/yr); and (h) NPP (gC/m2).


2561

energy variables are strongly influenced by land use/landcover types. Figure 10 shows that SatE is commonly gener-ated from the lowland plains on the east, only sporadicallyon the local depressions on the high hills and around thebending corner in the southwest, which as mentioned previ-ously is a semienclosed subbasin that hosts a chain of smalllakes. The pattern of SatE is very similar to the top 10 cmsoil moisture map, which shows the lowland areas beinggenerally wetter than upland. This pattern indicates thatgroundwater lateral flow may have transported a largequantity of water from the western hills to the eastern lowplains, over the entire spatial scale of the watershed. Thishypothesis is supported by the groundwater recharge map,which shows the hills as mainly recharging areas and thelow plains as discharging areas (Recharge< 0). On theother hand, sensible heat, temperature, and net primary pro-duction (NPP) maps are primarily controlled by land usetypes. Areas with high sensible heat correspond to theurban centers. This pattern is also obvious from the soiltemperature. Ground surfaces of urban areas are, on annualaverage, 2� to 4� hotter than forested areas due to the‘‘urban heat island’’ effect [Landsberg, 1981]. These valuesare similar to reported values (e.g., 4.3�C in Vukovich[1983] and 2.1�C in Turkoglu [2010]). Unsurprisingly, theregions with the lowest annual temperature are the sameareas with the highest latent heat flux. The agriculturalland, dominated by soybean, has quite high soil tempera-ture, but its sensible heat flux is not higher than the urbanarea. Compared to the forested areas, soybean provides lessshading and dissipates less latent heat due to the rather lowLAI and low canopy height of soybean (maximum LAI isaround 2). Compared to the urban areas, soybean has alarger resistance to sensible heat. The highest values oflatent heat are found in the forested or perennially pondedareas near streams in the high hills. These areas have amplemoisture supply due to local groundwater convergence.Following the same rationale, we would expect higherlatent heat to be found on the low plains. However, such anexpected pattern is not visually identifiable from the latentheat flux map. The latent heat figure shows generally higherlatent heat in the northern half of the basin with vegetatedland covers. The southern tip of the basin is urbanized and,therefore, does not have enough vegetation to transpirewater. Apparently, the influence of groundwater flow on ETis overshadowed by the land use and other spatial trends.

[36] Strong influences of landform and topography onrunoff, energy, and vegetation processes have also beennoted [Band et al., 1993; Ivanov et al., 2008; Jencso andMcGlynn, 2011; Kim et al., 1999; Rihani et al., 2010] andmodeled [Agnese et al., 2007; Bracken and Croke, 2007;Laio et al., 2009; Troch et al., 2003]. To get a better under-standing of the spatiotemporal distribution of runoff genera-tion and groundwater influences on vegetation in thiswatershed, we show in Figure 11 the fluxes and states alonga hillslope transect. The straight transect line begins on thecentral hill, nearly follows the elevation gradient and ends inthe lowland plain, as highlighted by the line segment A-B inFigure 2. To reduce noise and facilitate our interpretation,the vegetation types along the profile are all set to DBF insimulation S5 (Table 2), and the soil texture is set uniformlyto the most typical values. In Figure 11, time is on the x axis,and distance along the transect (abbreviated as distance) is

on the y axis, such that viewing from top of a panel to thebottom corresponds to going from uphill to downhill. Weshow weekly averaged plots of surface runoff, top 10 cmsoil moisture, groundwater head (H), and NPP. Since therunoff values vary across several orders of magnitude, weshow the data after applying a cube-root transformation. Aswe can see, the generation of runoff is highly concentrated,both in space and in time. Unsurprisingly, the pattern of run-off correlates closely with the soil moisture pattern. There isa thin horizontal line (near distance¼ 4 km) of runoff gener-ation that occurs sporadically in the year. At this location,transect A-B intersects a level 3 stream (the Stony Creek).Apart from this line, runoff occurs primarily on the lowland,which corroborates the findings from Figure 10. The divideat a distance of 15 km, where the hillslope descends to theflat plains, is clearly identifiable. As discussed by Kim et al.[1999] and Salvucci and Entekhabi [1995], this divide distin-guishes two distinct zones, the upper of which is dominatedby infiltration and moisture-limited evaporation while thelower is characterized by discharge and runoff production.The peaks of runoff generation occur only in two short timewindows, one in early May and another in December, whenthere is little vegetation activity. Like in other years, severalspring and winter storm events generate most of the runoffvolume. Very low volumes of runoff are generated insummer even in the lowland, despite a large storm event inJuly, which corresponds to a relatively ‘‘quiet’’ summer pe-riod in the hydrograph. Compared to runoff, groundwaterhead (H) varies much more smoothly, demonstrating spatio-temporal continuity. In summer, the contour lines of H bendupward, corresponding to the drydown of groundwater lev-els. From Figure 10h, it is clear that the availability of waterplays a lesser role than the seasonal energy input. NPP, bothuphill and downhill, responds primarily to seasonality. Theplants achieve maximum productivity in June and July, fol-lowing the maturity of the LAI and the high energy input.However, during these 2 months, the influence of moisture isalso visible. The lowland trees have approximately 15%higher NPP than uphill trees. Besides NPP magnitude,downhill trees also have longer period of high productivity.In June, there is a noticeable gap between the productivitypeaks. Reading from the precipitation and runoff plots, wesee that the timing of this gap corresponds to the hiatusbetween two large storm events. This period of lowered pro-ductivity is clearly due to moisture limitation. The gap isvery brief and shallow for the lowland trees, owing togroundwater subsidy.

[37] The rich patterns of different fluxes and states areapparently dominated by different processes. These inter-esting observations force us to ask the questions: How dothese processes produce, either in isolation or in concert,the observed spatiotemporal patterns? Since both ground-water and vegetation type seem to influence ET and NPP,which has a stronger influence? How relatively importantis each process on different hydrologic and vegetationprocesses? What is the role of soil water retention and un-saturated conductivity? We use hypothetical simulations toanswer these questions in the next section.

3.4. Relative Strengths of Controls

[38] In order to answer the questions raised at the end ofsection 3.3, we design a series of simulations to quantify


2562

the impacts of different factors on long-term annual ET andNPP as well as other variables (infiltration, runoff, transpi-ration, soil temperature, and soil moisture). For simplicity,we generically refer to the outcome variable as u. Detailsof these simulations and the objectives of the numericalexperiments are summarized in Table 2 and will be dis-cussed in the following sections. In this paper, we put ourfocus on four factors: land use types, topography-inducedgroundwater flow, soil water retention, and soil nitrogenlevels. Other factors such as slope aspect and micrometeor-ology may also be important for some sites [see Ivanov

et al., 2008], but they are outside the scope of this paper.We first characterize the effect of each factor and searchfor an appropriate surrogate index for the factor. Realizingthat each process is a complex combination of differentcontrols (e.g., for soil properties, soil conductivity, andwater retention characteristics all affect plant growth), weattempt to simplify our analysis by considering both soilcharacteristics and groundwater convergence as lumped‘‘factors.’’ Then, we evaluate the relative strength of eachcontrol in multivariate simulations using analysis of var-iance (ANOVA).

Figure 11. (a) Precipitation (mm/day). (b-e) are spatiotemporal plots of fluxes and states along the hill-slope transect A-B. The x axis is time and the y axis is distance from point A along the transect. Readingfrom top of the panel to the bottom is going from upland to lowland. (b) Runoff (mm/yr) ; (c) soil mois-ture content (m3/m3); (d) groundwater head (m); and (e) NPP (gC/m2/yr). The runoff values in Figure11b are cube-root transformed because the original values span several orders of magnitude and twopeaks overwhelm all other events. Elevation plot is placed to the left of the panels to show the topo-graphic descent of the hillslope.


2563

3.4.1. Single Factor Analysis and Process Indices[39] We refer to the baseline simulation as S0, which

most closely represents the realistic setting of the basin andis used to generate all results in the previous sections. Toevaluate the impact of land use types, we carry out a hypo-thetical simulation, referred to as S1, with randomly dis-tributed land uses, while all other factors are kept identicalto S0. In S1, five land use types (namely, temperate ever-green forest, temperate deciduous forest, grass, corn, andmedium-intensity urban), each assigned equal areas, arerandomly placed in the watershed to eliminate any spatialcorrelation with soil types, topography, and convergencelevels. The resulting long-term annual average ET is sum-marized by land use types and presented in the boxplot inFigure 12. The simulated ET are in general agreement withliterature reported values for DBF [e.g., Oishi et al., 2010],corn, and soybean [e.g., Suyker and Verma, 2009]. The ETfor corn is somewhat in the lower range of the reported irri-gated plants [e.g., Howell et al., 1997], because irrigation isnot simulated. As we can see, the interspecies variation ofET can be significant. Mean forest ET is higher than that ofcorn, which is higher than ET of grass. However, all vege-tated lands significantly have greater ET values than imper-

vious areas, supporting the spatial pattern observed inFigure 10. The results suggest that urbanization has a sig-nificant impact on ET fluxes and explains why its spatialpattern overwhelms that of topography and groundwaterconvergence. To further study the impact of other factors,we carry out the ensuing analysis in this section afterremoving the masking effect of urbanization.

[40] In simulation S2, the land use types are uniformlyset to deciduous broadleaf forest (DBF). We disablegroundwater lateral flow by setting the lateral conductivityof the aquifers to 10�14 (m/day). As a result, the cells arealmost isolated from each other, as in the original CLMmodel. Because other variables, including vegetation typeand lateral flow, are homogenized, we argue that the NPPobtained from S2 also measures the overall ability of soil(including the effect of soil water retention and thickness)to retain and supply water for tree growth. This effect isinfluenced not only by soil water retention characteristicsbut also by conductive properties and thickness. Since bothsoil retention and conductivities are heavily dependent ontexture, we may roughly refer to this physical factor as soiltexture. Therefore, the NPP values obtained from S2 can beviewed as a ‘‘soil water supply index’’ for NPP, orSWSI(NPP). More generally, we define SWSI(u,x) as thelong-term annual average values of variables u from S2 atspatial location x in the basin. For simplicity, x is omittedin the ensuing notations. u can be either ET, NPP, runoff,infiltration, or any other variable we choose. In the presentsimulation, SWSI(u) then represents how spatial variabilityof soil texture impact u. SWSI(NPP) are shown as a spatialmap in Figure 13a.

[41] An index for the groundwater flow process can beextracted from simulation S3, which has all settings identi-cal to S2 except that lateral groundwater flow is enabled.Therefore, the difference uS3� uS2 represents the net effectof lateral groundwater flow on variable u. We thereforedefine a ‘‘lateral flow index’’ for the variable u, or LatI(u),as the difference of long-term (5 years, 2005–2009) annualaverage u between S3 and S2. NPP S3 � NPP S2ð Þ=NPP s2 �100% is shown in the map in Figure 13b. As we can see,lateral groundwater flow subsidizes vegetation in thelowland plains under drought conditions and increases theirannual productivity by �5%. However, this subsidy comesat the cost of a 10–15%, even up to 30% in some cases,reduction of NPP on the highland hill areas. Earlier studies

Figure 12. Boxplot of land use influences on ET fromsimulation S1, where land use types are randomly placed inthe watershed and the correlation with spatial locations areremoved. The land use types are temperate needleleaf ever-green forest (ENF), temperate deciduous broadleaf forest(DBF), C3 grass (C3g), C3 general crop (C3c), corn, soy-bean (soybn), and impervious (imp.)

Table 2. List of Hypothetical Simulations

Simulation Procedures Purpose

S0 Use the baseline parameters To serve as the baseline scenarioS1 Based on S0, land uses are randomly distributed in the

watershedTo show the intraclass and interclass variability of ET

S2 Based on S0, land use uniformly set to DBF, lateralgroundwater disabled

To show the influence of soil characteristics on plantgrowth (in terms of water supply) and also to serve as a

basis for comparison with S3S3 Based on S0, land use uniformly set to DBF Together with S2, to show the influence of lateral

groundwater flowS4 Based on S0, land use randomly selected from vegetated

classes: temperate evergreen tree, broadleafdeciduous tree, C3 grass, and corn

To assess the relative significance of the controllingfactors with effect of urbanization silenced.

S5 Same simulation as S0, however, along the topographicgradient (line A-B), land use is set to D BF, and soil is

set to the dominant soil type along the paths.

To demonstrate the influence of spatiotemporaldistribution of runoff, soil moisture, groundwater head,

and NPP.


2564

[Ivanov et al., 2008, 2010] that focused on arid, moisture-limited environments showed stronger influence of lateralgroundwater redistribution on vegetation growth. Ourresults indicate that, even for an energy-limited region asthe present watershed, groundwater may still play a notice-able role in regulating plant growth, by subsidizing lowlandplants while sacrificing upland productivity. The ground-water influence, in our case, also seems to impact plants onthe highland much more significantly than those on thelowland plains. Interestingly, the subsidy seems higher im-mediately adjacent to the foot of the hills and reduces aswe go deeper into the plains. This is because, on the flatplains, the trees at the foot of the hill have the best accessto the lateral groundwater discharge that comes from uphill.On a side note, one would expect LatI to be correlated totopographic index (TopoI) [Beven and Kirkby, 1979], a‘‘wetness’’ index that aims to provide a simple metric toestimate the likelihood of saturation. However, as shown inthe (LatI-TopoI) plot in Figure 14b, such a correlation isnot obvious (plotting TopoI against ET yields a similarplot). Roughly, a rectangle could be drawn to enclose thescattered dots. However, TopoI is unable to explain a largeportion of the variability, which might include the hetero-geneity of aquifer properties, foothill dynamics, and theparticular topographic and watershed construct (concave ordiverging, etc.), among other factors. It is known that thetopographic index may not work well in flat terrains. There-

fore, for the purpose of comparing their relative influences,we choose SWSI and LatI to represent, respectively, soilproperties and groundwater flow dynamics.

[42] Although the representation of the N cycle in CLMis still too crude compared to reality, it does allow us tomake some initial assessment of the order of magnitude ofthe influence as compared to other factors. Over a longterm, a dynamic balance is established between nitrogen inthe plant, litter, SOM and mineral pools, BNF, deposition,leaching, and denitrification. Random perturbation of the Npools will lead to a rebalancing of the system and a gradualregression to the equilibrium. In order to express the N var-iability while approximately maintaining such a balance,we set the BNF rates using equation (1) [Cleveland et al.,1999; Luo et al., 2006; Oleson et al., 2010]:

BNF ¼ �� 1:8 1� e�0:003NPP� �

; (1)

where BNF is in gN/m2/yr, NPP is the annual net primaryproduction, and a is an adjustment factor whose values arekept at these hypothetical levels : � 2 0:2; 0:6; 0:8;1; 1:2; 1:4; 1:8. We choose this range as it resembles therange of the uncertainty with natural BNF rate estimates[Galloway et al., 2004]. In the original CLM, a is alwaysequal to 1.0. In our relative control analysis, the model isspun up over 4000 years by looping through the climaticforcing data between 2001 and 2009, as in Thornton andRosenbloom [2005], but with different a values as givenabove. The final C and N states at steady state differ greatlybetween different assumed BNF rates. Table 3 shows sometypical maximum LAI values, end-of-year total SOM car-bon/nitrogen states as a result of the different BNF rates.While the alteration of the BNF rates changes not only Npools but also C pools, we refer to this factor as nitrogenlevels, to reflect the fact that nitrogen is the limiting factor.We note that although we perturbed the BNF rates function,we do not posit any causal or ordinal relationship betweenBNF and NPP change. This is just a way for us to expressthe variability of nitrogen availability and propagate itthrough the ecosystem, while maintaining the balancebetween different N pools. Of course, we may also choose

Figure 13. Spatial distributions of (a) NPP from S2, rep-resenting the soil’s ability to provide water to plants and(b) NPP percentage change from S2 to S3 (the difference istermed LatI), showing the net effect of lateral groundwaterflow in subsidizing lowland trees while reducing availablemoisture for upland trees.

Figure 14. LatI(NPP) (which is NPP of S3 minus NPP ofS2) as a function of topographic index. As we can see, thetopographic index is not a very good predictor for NPP inthis watershed.


2565

nitrogen deposition rates as the knob to perturb, whichwould generate similar effects. The C/N pools at the equi-librium corresponding to different a values are recorded tobe used as initial states in the following combinedsimulations.3.4.2. Combined Analysis Using ANOVA

[43] Using ANOVA, we can quantitatively assess the rel-ative strengths of the controlling processes by approxi-mately estimating the variance of vegetation growthattributable to each one of the factors. In simulation S4,baseline soil water retention properties and aquifer conduc-tivities are used. To mask the dominant role of urbaniza-tion, the land cover in S4 are limited to evergreen forest(ENF), DBF, C3 grass, and soybean, and these classes arerandomly placed in the watershed. The nitrogen level ofeach cell is randomly coded between 1 and 7. For eachcoded level, the initial C/N states are looked up from thelong term spin-up results with the corresponding BNF ratedescribed in the last paragraph.

[44] A four-way ANOVA test (assessed at the 1% confi-dence level) with interactions is used to partition the totalvariance of u in S4 into the individual contributions. Thefour predictors are land use types (PFT), soil (SWSI(u)),lateral GW flow (LatI(u)), and nitrogen levels (N level).The ANOVA results are presented in Tables 4 and 5 andgraphically in Figure 15 for different u. We examine sevendifferent u’s, which are the long-term annual averages ofET, NPP, runoff (Rf), infiltration (Inf), top 10 cm soil tem-perature (t10cm), and top 10 cm soil moisture (�10cm). Thecross terms are generally small compared to the main fac-tors. For simpler presentation, they are combined as an‘‘interactions’’ term in Figure 15. The ‘‘error’’ term, whichrepresents the variance that cannot be explained by the

main factors and their interactions, is also shown. The par-titioned variances sum up to 1 for all u’s.

[45] It is obvious from the table that all factors are statis-tically significant for all variables, but their relative impor-tance shifts for different u, revealing the controllingprocesses. Groundwater lateral flow dynamics clearly dom-inate hydrologic fluxes like runoff and infiltration (but notET), with LatI weighing more than 70% and 76%, respec-tively. The soil textural properties (SWSI) rank next toLatI, explaining 16% and 11% of the variance, respec-tively. LatI’s dominance over SWSI for Rf and Inf can berelated to the fact that saturation excess is the primarystreamflow generation mechanism in the basin, and thestate of saturation in space and time is strongly controlledby groundwater dynamics. NPP is primarily controlled bynitrogen levels, PFT, and their interactions. The cross term(N level � PFT) occupies more than 97% of the interac-tions. The results highlight N as a limiting factor of vegeta-tion growth in the system. The top 10 cm soil temperatureis almost completely driven by land use types, even afterurban land use is excluded from S4. The surface soil tem-perature t10cm reflects the optical and thermal properties ofthe canopy structure. Although CLM4.0 does consider thedifference in thermal properties between water and soil(PAWSþCLM does not track heat transfer by water), nei-ther LatI nor SWSI seem to be have any noticeable impact.

[46] The most interesting variable is ET, for which all ofLatI, SWSI, PFT, and N levels and their interactions areimportant. This is attributed to the colimitation by soilwater, energy and vegetation, which is unique to this humidcontinental climate where no single factor dominates. Theecosystem controls (N level : 23%, PFT: 38%, their interac-tions: 8%) seem to be stronger than water controls (SWSI:

Table 3. Different Ecosystem States (Maximum Value in a 9 Year Period) for DBF at Steady State as a Result of Different BNF Ratesa

a 0.2 0.6 0.8 1.0 1.2 1.4 1.8

LAI 4.0 4.8 5.1 5.5 5.8 6.2 6.8Canopy top height (m) 18.4 20.5 21.5 22.3 23.0 23.6 24.7Total soil organic matter carbon (gC/m2) 5,930.2 7440.3 8,195.2 8,913.5 9,535.1 10,154.2 11,234.2Total soil organic matter Nitrogen (gN/m2) 591.3 741.9 817.1 888.8 950.8 1,012.5 1,120.2Total ecosystem carbon (gC/m2) 13,770.9 18,275.9 20,562.3 22,731.2 24,645.0 26,519.8 29,805.8Total ecosystem nitrogen (gN/m2) 623.5 783.0 862.7 938.6 1,004.3 1,069.7 1,183.9

aBNF ¼1:8a 1� e�0:003NPP anð Þ where NPPan is the annual NPP and a is the adjustment factor.

Table 4. Four-Way ANOVA of N Level/Land Use Type/SoilTexture/Groundwater Flow on u¼ET

SourceSum ofSquares

Percent ofVariance

Explained (%) F Prob > F

N level 4,835,447 23.53 1,783 0PFT 7,831,229 38.11 5,775 0SWSI 1,988,126 9.67 4,398 0LatI 3,242,024 15.78 7,172 0N level � PFT 1,100,465 5.35 135 0N level � SWSI 131,682 0.64 49 0N level � LatI 100,415 0.49 37 0PFT � SWSI 95,569 0.47 70 0PFT � LatI 158,611 0.77 117 0SWSI � LatI 10,647 0.05 24 1.30E-06Error 1,056,450 5.14Total 20,550,664 100.00

Table 5. Four-Way ANOVA of N Level/Land Use Type/SoilTexture/Groundwater Flow on u¼NPP

SourceSum ofSquares

Percent ofVariance

Explained (%) F Prob > F

N level 84,307,345 72.47 25,590 0PFT 2,594,772 2.23 1,575 0SWSI 742,608 0.64 1,352 0LatI 253,061 0.22 461 0N level � PFT 24,127,980 20.74 2,441 0N level � SWSI 2,326,108 2.00 706 0N level � LatI 632,842 0.54 192 0PFT � SWSI 62,980 0.05 38 0PFT � LatI 10,551 0.01 6 2.56E-04SWSI � LatI 5 0.00 0 9.27E-01Error 1,281,601 1.10Total 116,339,853 100.00


2566

10% and LatI: 15%). Soil nitrogen level (as a result of eco-system spin-up to equilibrium using different BNF rates)plays an important role in ET, confirming previous obser-vations of positive correlation between BNF and ET. As isclear from Figure 15, this influence is mainly exertedthrough its control on transpiration (Tp). It can then beinferred that groundwater and soil texture are much moredominant controls on soil evaporation. The influence of Nlevels also propagates to hydrologic fluxes and tempera-ture, although not in a strong fashion. The findings suggestthat in order to accurately predict spatiotemporal distribu-tions of ET (or latent heat flux), the ecosystem dynamicsincluding nitrogen availability must be carefully considered.

[47] In contrast, LatI and SWSI have comparable signifi-cance for �10cm and each explain about 40% of the var-iance. As a central variable for hydrologic andbiogeochemical processes, soil moisture has been the mostintensively studied in the literature among the variables an-alyzed in this paper. Interestingly, our results have beencorroborated by analysis of field data [Famiglietti et al.,1998; Joshi and Mohanty, 2010; Pan and Wang, 2009;Wilson et al., 2004], where soil and topography are foundto be of similar importance (although soil/vegetationimpacts are lumped in Wilson et al. [2004], and it is notclear how much variance they each explain). Therefore, forhydrologic and biogeochemical processes, which dependon soil moisture in a strongly nonlinear fashion, it is impor-tant to consider both local soil textural properties andregional groundwater flow. Qiu et al. [2001] report a largercontrol of mean soil moisture by land use. The actualcontrols of vegetation on soil moisture are expected to behigher than explained in the model. Further study should

consider more diversified vegetation and the effects ofhydraulic lift by vegetation or macropores created by vege-tation roots. Our study intentionally breaks the correlationthat naturally exists between topography and vegetationtypes [e.g., Brown, 1994]. We also note a large error for�10cm, indicating that the complex dynamics of moisturecannot be explained by simple linear models of the varia-bles. The complex controls of soil moisture have beenreported by Zhu and Lin [2011], who conclude that terrainattributes are more important in wetter places among otherdynamics.

[48] It must be noted that, for NPP, although LatI andSWSI appear unimportant compared to vegetation typesand soil nitrogen content, both have significant influenceonce we hold PFT and N levels constant across the basin,as we have shown in S3.3.4.3. Robustness of the Analysis

[49] A reasonable question to ask is whether the relativeimportance of different processes will stay the same whenwe perform the same exercise with a different model, a dif-ferent spatial resolution or in another region where climaticinput and physical parameters (e.g., aquifer transmissiv-ities) are different. Intuitively, the results should changewhen physical parameters are altered since, if we imagine abasin with very low aquifer transmissivities, the influenceof groundwater flow dynamics should be very low.Although the effect of different model formulations shouldalso be examined in the future, in this paper, we assess theeffect of the parameters and spatial resolution. To quantita-tively assess this impact, we redo the ANOVA analysis insection 3.4.2 with exactly one of the following changesapplied: (a) the conductivity of the unconfined aquifer

Figure 15. ANOVA analysis results show the relative strengths controls of physical processes onhydrologic and vegetation dynamics (Rf, runoff ; Inf, infiltration; ET, evapotranspiration; Tp, transpira-tion; NPP, net primary productivity; t10cm, top 10 cm soil temperature; �10cm, top 10 cm soil moisture;Interactions, for simplicity, the cross terms between different factors are combined into one interactionterm). ‘‘Original’’ stands for ANOVA test carried out without global parameter changes described in sec-tion 3.4.3. In ‘‘K � 0.1’’ ANOVA test, all simulations (S2, S3, and S4) have their K scaled by 0.1. Simi-larly, 0.1 is added to N in all simulations in ‘‘Nþ 0.1’’ ANOVA test. ‘‘Double resolution’’ means thespatial step size is halved (number of cells quadrapled). (a) Original, (b) K � 0.1, (c) double resolution,(d) Nþ 0.1.


2567

everywhere in the basin is reduced by a factor of 10. There-fore, the basin average unconfined aquifer K becomes1.3 m/day; (b) the van Genuchten parameter n is increasedby 0.1, thereby increasing the nonlinearity of the soil waterretention and unsaturated conductivities ; (c) the horizontalcell size is reduced in half (resolution is doubled), therebyquadrupling the number of cells in the basin.

[50] The results are shown in Figure 15. Indeed, the var-iances attributed to each factor do shift after making thechanges (Figure 15a). The influence of LatI reduces for allvariables, implying a less dominant groundwater system.The relative roles of other factors, especially SWSI,increased accordingly to make up for the drop in LatI. Thismeans the relative importance of groundwater dynamicswill shift when we move to a geologic configuration thatextends beyond the current range of variability. Withchange (b), however, the changes are relatively minor, withSWSI explaining slightly more variance in Rf and Inf. Theshifts in the relative importance due to a change in therange of variability have also been previously noted byfield data analysis [Zhu and Lin, 2011]. With the help of anumerical model, however, the relative importance of con-trols at each site can be analyzed prior to extensive datacollection.

[51] The spatial resolution has a relatively weak influ-ence on the results of the analysis. For ET, Rf, and Inf, LatInow plays a more important role when resolution isdoubled, scoring 20%, 74%, and 87%, respectively. This isdue to the better resolution of the effects of topography rel-ative to the original simulation. For t10cm and �10cm, how-ever, SWSI increased slightly, presumably due to a largerrange of SWSI as higher resolution allows more distinctsoil classes to be simulated. However, the general trendstays the same as in the lower resolution simulations.

4. Discussion and Conclusions

[52] The knowledge and the mechanistic understandingobtained in this study may help guide future modeling anddata-gathering decisions. For example, many previousefforts focused on downscaling satellite soil moistureobservations [e.g., Busch et al., 2012; Mascaro et al.,2010; Pellenq et al., 2003], and our results indicate thatboth topography and soil texture should be environmentalpredictors in such downscaling methods as they are impor-tant controls for soil moisture. For another example, forgross primary production (GPP/NPP) estimation, at leastfor the humid continental climatic region examined here,plant type and nutrient availability far outweigh ground-water dynamics and soil water supply properties in terms oftheir relative importance. It is therefore important to obtainsoils data sets with better carbon/nitrogen storage estima-tions. For quantification of terrestrial-atmospheric vapor/heat exchange, groundwater dynamics, soil properties, andecosystem processes are all important. Because both short-ranged and long-ranged patterns of groundwater flow exist,it is important to consider regional groundwater dynamicsand subsurface processes in land surface models. Thismeans that for large-scale simulations that aim at studyingterrestrial-atmospheric vapor/energy exchange, many sim-plified cell-based hydrologic formulations (such as theTOPMODEL-based formulations in original CLM and

Noah LSM) are likely to produce unreliable estimates.Although similar conclusions have been drawn in the litera-ture [Anyah et al., 2008; Fan et al., 2007; Maxwell andKollet, 2008a], our study is unique in this climatic region.

[53] In spite of the limitations of the current nitrogencycling schemes in CLM4.0, the results clearly indicate thatthe nitrogen availability heavily influences not only vegeta-tion growth but also ET, through which other hydrologicfluxes are affected. Conventional hydrologic models oftendo not consider the nutrient dynamics mechanistically. Forassessment of climate change on regional hydrology andvegetation, nitrogen dynamics should be considered. Theresults in this paper suggest that we need to take an integra-tive view of the ecosystem-hydrology interactions in orderto make better predictions. It is important for hydrologiststo continue expanding the domain of sciences in hydrologicsimulations to identify previously undiscovered but poten-tially important linkages, echoing the call for more cross-interdisciplinary integration [Wagener et al., 2010] andencouraging grass-roots community collaborations.

[54] For many variables studied in Figure 15 (e.g., ET,Tp, NPP, and t10cm), the error terms account for less than10% of the total variance. Consequently, it is possible topredict these variables (at least their annual averages) witha fair degree of accuracy by just using simple linear modelsof the main processes (land use, nitrogen, groundwaterflow, and soil water retention) and their interactions. How-ever, the variables with higher percentage of error (e.g.,�10cm) will need to be addressed by more complex modelsdetailing the space-time interactions of factors.

[55] As with other similar studies, the conclusions in thepaper are somewhat tied to the model employed. The abil-ity of a numerical model to represent linkages and interac-tions depends on its formulations. Considering the largeuncertainties associated with descriptions of the nitrogencycle in CLM, the numerical value of its relative impor-tance will likely shift when improved N cycling schemesare employed. Given that a large range of BNF rates(altered by a) are used to get carbon nitrogen states, thetrue influence of nitrogen is likely smaller than estimated.Future simulations supported by novel observations will becrucial for confirming the results. PAWSþCLM considerswater to move only vertically in the unsaturated zone. As aresult, the model does not produce accurate results whensignificant interflow exists. The model currently does nottrack nutrient transport within groundwater and surfacewater. Also, the analysis in the paper does not include theeffects of microtopography, micrometeorological condi-tions, and aspect of the slope. In the future, it would beinteresting to compare results from other models using theapproach described here. The effect of very deep soils onthe coupling algorithm in PAWS seems to be mild giventhe good match with observed groundwater heads and well-captured baseflow in the hydrograph. In the future, veryhigh resolution simulations will likely provide more insighton the effect of spatial resolution. Scale-aware or multi-scale simulation approaches are called for to help with theissue of spatial scaling.

[56] To summarize, we have described the application ofa novel process-based hydrologic model PAWSþCLM to awatershed in the humid continental climate region of theU.S. Midwest to understand the relative importance of


2568

vegetation type, nitrogen availability, groundwater dynam-ics, and soil texture. The model shows good comparisonwith various observation data sets. The background hydrol-ogy of the basin is revealed as one limited by energy andwater at different times during the annual cycle, with satu-ration excess being the dominant runoff generating mecha-nism. Groundwater plays an important role in thewatershed, controlling saturation excess and lowland exfil-tration. Through a series of hypothetical simulations basedon the model, we use ANOVA to quantify the influenceand relative importance of the controlling processes. Theresults confirm the basin, which has an average unconfinedaquifer conductivity of 13 m/day, as a groundwater-domi-nated system. Groundwater flow has been found to be themajor control on runoff and infiltration, explaining morethan 70% of the variance. Soil texture ranks next. Vegeta-tion type and nitrogen levels are found to determine NPP,top 10 cm soil temperature, and transpiration. Interestingly,all factors are found to significantly control ET, whichserves as the nexus for ecosystem-hydrology interactions.Nitrogen is shown to be a major control of ecosystem varia-bles, but it also significantly controls transpiration, throughwhich it influences other hydrologic fluxes. While keepingin mind the limitations with the CN cycling in CLM, theresults indicate that hydrologists need to continue expand-ing the domains of sciences in hydrologic simulations todiscover important linkages. The analysis method used inthe paper is shown to be relatively robust with respect tochanges in parameter values and spatial resolution, butwhen we move to a region whose parameters extendbeyond the current range of variability, the results areexpected to shift.

Appendix A: Lowland-Storage Module

[57] Here we describe a new enhancement to the model.Wetlands are an important land cover type that exertsstrong influence on hydrology, ecosystem functioning, andwater quality. A flexible lowland-storage module is addedto the original PAWS model to keep track of the exchangesbetween wetlands, overland flow, and groundwater. Thewetland is conceptualized as a lowland-storage compart-ment of the overland flow domain, as shown in Figure 1.To accommodate the wetland compartment, the mass bal-ance equation for the flow domain is modified as follows:

@hf

@t¼ fwKw

H � zw þ hf =fw

� ��zw

� Ew þ Fg �r hof uð Þ; (A1)

in which hf is the flow depth in the flow domain, H is thegroundwater head, Ew is the evaporation from lowland stor-age, fw is the areal fraction of lowland storage, Fg is runofffrom the ponding domain, u is overland flow velocity vector,Kw is the bottom conductivity, zw is the bottom elevation ofthe lowland storage, �zw is the maximum of zw�H or thick-ness of bed material layer, and hof is the overland flow depthin excess of the lowland-storage capacity. Unit of the equa-tion is m/day. We solve equation (A1) in two fractional steps.The first one is an implicit step that calculates the ground-water exchange and obtains intermediate states h�f . The sec-ond step takes hof ¼ max 0; h�f � dw

� �, where dw is the

bucket depth of the depression storage, and sends result to

the 2-D diffusive wave overland flow solver as described inSP10 to solve for the convective term, rhof u. The exchangebetween groundwater and lowland storage becomes eitherexfiltration (Ex) or infiltration (Inf), depending on the direc-tion. Equation (A1) can also be applied to describe transientshallow concentrated flows, paddy fields, or potholes. Inaddition, the lowland storages in different locations are con-nected via overland flow paths. When flooding occurs, riverwater may flood the overland flow domain, and in turn, fillthe lowland-storage compartments. The threshold for backfillto happen is termed as hback, fw, and dw, all of which can beinferred from analyzing land cover maps and fine-resolutiondigital elevation model (DEM).

[58] Acknowledgments. J.N. and M.S.P. are supported by the NOAACenter of Excellence for Great Lakes and Human Health. We thank SamLevis, NCAR, for valuable advice related to CLM and Ben Ong, Institutefor Cyber-Enabled Research (iCER) at MSU for help with high-perform-ance computing. We thank three anonymous reviewers whose constructiveand detailed comments have greatly helped in improving the manuscript.

ReferencesAgnese, C., G. Baiamonte, and C. Corrao (2007), Overland flow generation

on hillslopes of complex topography: Analytical solutions, Hydrol. Pro-cess., 21(10), 1308–1317, doi:10.1002/hyp.6354.

Anyah, R. O., C. P. Weaver, G. Miguez-Macho, Y. Fan, and A. Robock(2008), Incorporating water table dynamics in climate modeling: 3.Simulated groundwater influence on coupled land-atmosphere variabili-ty, J. Geophys. Res., 113, D07103, doi:10.1029/2007JD009087.

Arora, V. (2002), Modeling vegetation as a dynamic component in soil-vegetation-atmosphere transfer schemes and hydrological models, Rev.Geophys., 40(2), 3-1–3-24, doi:10.1029/2001RG000103.

Band, L. E., P. Patterson, R. Nemani, and S. W. Running (1993), Forestecosystem processes at the watershed scale: Incorporating hillslope hy-drology, Agric. For. Meteorol., 63(1-2), 93–126, doi:10.1016/0168–1923(93)90024-c.

Band, L. E., C. L. Tague, P. Groffman, and K. Belt (2001), Forest ecosystemprocesses at the watershed scale: Hydrological and ecological controls ofnitrogen export, Hydrol. Process., 15, 2013–2028, doi:10.1002/hyp.253.

Berg, B., and R. Laskowski (2005), Litter Decomposition: A Guide to Car-bon and Nutrient Turnover, Adv. Ecol. Res. 38, p. 448, Academic Press,Waltham, MA.

Beven, K. (2002), Towards an alternative blueprint for a physically baseddigitally simulated hydrologic response modelling system, Hydrol. Pro-cess., 16(2), 189–206.

Beven, K., and M. Kirkby (1979), A physically based, variable contributingarea model of basin hydrology, Hydrol. Sci. Bull., 24(11), 43–69.

Bonan, G. B., and S. Levis (2010), Quantifying carbon-nitrogen feedbacksin the Community Land Model (CLM4), Geophys. Res. Lett., 37,L07401, doi:10.1029/2010GL042430.

Bracken, L. J., and J. Croke (2007), The concept of hydrological connectiv-ity and its contribution to understanding runoff-dominated geomorphicsystems, Hydrol. Process., 21(13), 1749–1763, doi:10.1002/hyp.6313.

Brown, D. G. (1994), Predicting vegetation types at treeline using topogra-phy and biophysical disturbance variables, J. Veg. Sci., 5(5), 641–656.

Burke, I. C., W. K. Lauenroth, and W. J. Parton (1997), Regional and tem-poral variation in net primary production and nitrogen mineralization ingrasslands, Ecology, 78(5), 1330–1340, doi:10.1890/0012–9658.

Burt, T. P., and G. Pinay (2005), Linking hydrology and biogeochemistry incomplex landscapes, Prog. Phys. Geogr., 29(3), 297–316, doi:10.1191/0309133305pp450ra.

Busch, F. a., J. D. Niemann, and M. Coleman (2012), Evaluation of an em-pirical orthogonal function-based method to downscale soil moisture pat-terns based on topographical attributes, Hydrol. Process., 26, 2696–2709, doi:10.1002/hyp.8363.

Celia, M. A., E. T. Bouloutas, and R. L. Zarba (1990), A general nass-con-servative numerical-solution for the unsaturated flow equation, WaterResour. Res., 26(7), 1483–1496.

Chorover, J., et al. (2011), How water, carbon, and energy drive criticalzone evolution: The Jemez–Santa Catalina Critical Zone Observatory,Vadose Zone J., 10, 884, doi:10.2136/vzj2010.0132.


2569

Cleveland, C. C., et al. (1999), Global patterns of terrestrial biological nitro-gen (N-2) fixation in natural ecosystems, Global Biogeochem. Cycles,13(2), 623–645, doi:10.1029/1999GB900014.

Collatz, G. J., M. Ribas-Carbo, and J. A. Berry (1992), Coupled photosyn-thesis-stomatal conductance model for leaves of C4 plants, Aust. J. PlantPhysiol., 19(5), 519–538.

Collins, W. D., et al. (2006), The Community Climate System Model ver-sion 3 (CCSM3), J. Clim., 19(11), 2122–2143.

D’Odorico, P., F. Laio, A. Porporato, and I. Rodriguez Iturbe (2003),Hydrologic controls on soil carbon and nitrogen cycles. II. A case study,Adv. Water Resour., 26, 59–70.

Dai, Y. J., R. E. Dickinson, and Y. P. Wang (2004), A two-big-leaf modelfor canopy temperature, photosynthesis, and stomatal conductance,J. Clim., 17(12), 2281–2299.

Dentener, F., et al. (2006), Nitrogen and sulfur deposition on regional andglobal scales: A multimodel evaluation, Global Biogeochem. Cycles, 20,GB4003, doi:10.1029/2005GB002672.

Dickinson, R. E. (1983), Land surface processes and climate surface albe-dos and energy-balance, Adv. Geophys., 25, 305–353.

Dickinson, R. E., K. W. Oleson, G. Bonan, F. Hoffman, P. Thornton, M.Vertenstein, Z. L. Yang, and X. B. Zeng (2006), The Community LandModel and its climate statistics as a component of the Community Cli-mate System Model, J. Clim., 19(11), 2302–2324.

Dingman, S. L. (2008), Physical Hydrology, 2nd ed., Waveland Press,Long Grove, IL.

Enviro-weather (2010), Enviro-weather Automated Weather Station Net-work (formerly known as MAWN). [Available at http://www.agwea-ther.geo.msu.edu/mawn/.].

Famiglietti, J. S., J. W. Rudnicki, and M. Rodell (1998), Variability in sur-face moisture content along a hillslope transect: Rattlesnake Hill, Texas,J. Hydrol., 210(1-4), 259–281, doi:10.1016/s0022-1694(98)00187-5.

Fan, Y., G. Miguez-Macho, C. P. Weaver, R. Walko, and A. Robock(2007), Incorporating water table dynamics in climate modeling: 1.Water table observations and equilibrium water table simulations,J. Geophys. Res., 112, D10125, doi:10.1029/2006JD008111.

Fang, C., P. Smith, J. U. Smith, and J. B. Moncrieff (2005), Incorporatingmicroorganisms as decomposers into models to simulate soil organicmatter decomposition, Geoderma, 129, 139–146, doi:10.1016/j.geoderma.2004.12.038.

Farquhar, G. D., S. V. Caemmerer, and J. A. Berry (1980), A biochemical-model of photosynthetic CO2 assimilation in leaves of C-3 species,Planta, 149(1), 78–90.

Fatichi, S., V. Y. Ivanov, and E. Caporali (2012), A mechanistic ecohydro-logical model to investigate complex interactions in cold and warmwater-controlled environments: 1. Theoretical framework and plot-scaleanalysis, J. Adv. Model. Earth Syst., 4, M05002, doi:10.1029/2011MS000086.

Ferguson, I. M., and R. M. Maxwell (2010), Role of groundwater in water-shed response and land surface feedbacks under climate change, WaterResour. Res., 46, W00F02, doi:10.1029/2009WR008616.

Galloway, J. N., et al. (2004), Nitrogen cycles: Past, present, and future,Biogeochemistry, 70(2), 153–226, doi:10.1007/s10533-004-0370-0.

Galloway, J. N., A. R. Townsend, J. W. Erisman, M. Bekunda, Z. C. Cai, J.R. Freney, L. A. Martinelli, S. P. Seitzinger, and M. A. Sutton (2008),Transformation of the nitrogen cycle: Recent trends, questions, and poten-tial solutions, Science, 320(5878), 889–892, doi:10.1126/science.1136674.

Global Soil Data Task (2000), Global soil data products CD-ROM (IGBP-DIS). International Geosphere-Biosphere Programme-Data and Informa-tion System. [Available at http://daac.ornl.gov/SOILS/guides/igbp.html,accessed 22 Nov. 2012.].

Goderniaux, P., S. Brouyere, H. J. Fowler, S. Blenkinsop, R. Therrien, P.Orban, and A. Dassargues (2009), Large scale surface-subsurface hydro-logical model to assess climate change impacts on groundwater reserves,J. Hydrol., 373(1-2), 122–138, doi:10.1016/j.jhydrol.2009.04.017.

Gunduz, O., and M. M. Aral (2005), River networks and groundwater flow:A simultaneous solution of a coupled system, J. Hydrol., 301(1-4), 216–234, doi:10.1016/j.jhydrol.2004.06.034.

Groundwater Inventory and Mapping Project (GWIM) (2006), State ofMichigan Public Act 148 Groundwater Inventory and Mapping Project(GWIM), IWR, Michigan State University, Technical Report. Availableat: http://gwmap.rsgis.msu.edu/ accessed on 22 November, 2012.

Hessl, A. E., C. Milesi, M. A. White, D. L. Peterson, and R. E. Keane(2004), Ecophysiological parameters for Pacific Northwest trees. Rep.PNW-GTR-618., U.S. Dep. of Agric., For. Serv., Pacific Northwest Res.Stn., 14 pp., Portland, Oreg.

Howell, T. A., J. L. Steiner, A. D. Schneider, S. R. Evett, and J. A. Tolk(1997), Seasonal and maximum daily evapotranspiration of irrigatedwinter wheat, sorghum, and corn—Southern High Plains, Trans. ASAE,40(3), 623–634.

Hwang, T., L. E. Band, J. M. Vose, and C. Tague (2012), Ecosystem proc-esses at the watershed scale: Hydrologic vegetation gradient as an indi-cator for lateral hydrologic connectivity of headwater catchments, WaterResour. Res., 48, W06514, doi:10.1029/2011WR011301.

Ivanov, V. Y. (2006), Effects of dynamic vegetation and topography onhydrological processes in semi-arid areas, Ph.D. dissertation, 453 pp.,Mass. Inst. of Technol., Boston, Mass.

Ivanov, V. Y., R. L. Bras, and E. R. Vivoni (2008), Vegetation-hydrol-ogy dynamics in complex terrain of semiarid areas: 2. Energy-watercontrols of vegetation spatiotemporal dynamics and topographicniches of favorability, Water Resour. Res., 44, W03430, doi:10.1029/2006WR005595.

Ivanov, V. Y., S. Fatichi, G. D. Jenerette, J. F. Espeleta, P. A. Troch, and T.E. Huxman (2010), Hysteresis of soil moisture spatial heterogeneity andthe ‘‘homogenizing’’ effect of vegetation, Water Resour. Res., 46,W09521, doi:10.1029/2009WR008611.

Jencso, K. G., and B. L. McGlynn (2011), Hierarchical controls on runoffgeneration: Topographically driven hydrologic connectivity, geology,and vegetation, Water Resour. Res., 47, W11527, doi:10.1029/2011WR010666.

Jia, Y. W. (1998), Modeling infiltration into a multi-layered soil during anunsteady rain, J. Hydrosci. Hydraul. Eng., 16(2), 1–10.

Joshi, C., and B. P. Mohanty (2010), Physical controls of near-surface soilmoisture across varying spatial scales in an agricultural landscape duringSMEX02, Water Resour. Res., 46, W12503, doi:10.1029/2010WR009152.

Kim, C. P., G. D. Salvucci, and D. Entekhabi (1999), Groundwater-surfacewater interaction and the climatic spatial patterns of hillslope hydrologi-cal response, Hydrol. Earth Syst. Sci., 3(3), 375–384.

Knyazikhin, Y., et al. (1999), MODIS leaf area index (LAI) and fraction ofphotosynthetically active radiation absorbed by vegetation (FPAR) prod-uct (MOD15) algorithm theoretical basis document [Available at http://eospso.gsfc.nasa.gov/atbd/modistables.html.].

Kollet, S. J., and R. M. Maxwell (2008), Capturing the influence of ground-water dynamics on land surface processes using an integrated, distributedwatershed model, Water Resour. Res., 44, W02402, doi:10.1029/2007WR006004.

Kondo, J., and S. Ishida (1997), Sensible heat flux from the earth’s surfaceunder natural convective conditions, J. Atmos. Sci., 54(4), 498–509.

Kundu, P. K., and I. M. Cohen (2010), Fluid Mechanics, 4th ed., AcademicPress, Waltham, MA.

Lai, C. T., and G. Katul (2000), The dynamic role of root-water uptake in cou-pling potential to actual transpiration, Adv. Water Resour., 23(4), 427–439.

Laio, F., S. Tamea, L. Ridolfi, P. D’Odorico, and I. Rodriguez-Iturbe(2009), Ecohydrology of groundwater-dependent ecosystems: 1. Sto-chastic water table dynamics, Water Resour. Res., 45, W05419,doi:10.1029/2008WR007292.

Landsberg, H. E. (1981), The Urban Climate, Academic Press, Waltham,MA.

Lawrence, D. M., et al. (2011), Parameterization improvements and func-tional and structural advances in version 4 of the Community LandModel, J. Adv. Model. Earth Syst., 3, M03001, doi:10.1029/2011MS000045, 27 pp.

Lohse, K. a., P. D. Brooks, J. C. McIntosh, T. Meixner, and T. E. Huxman(2009), Interactions between biogeochemistry and hydrologic systems,Annu. Rev. Environ. Resour., 34, 65–96, doi:10.1146/annurev.environ.33.031207.111141.

Luo, Y. Q., D. F. Hui, and D. Q. Zhang (2006), Elevated CO2 stimulates netaccumulations of carbon and nitrogen in land ecosystems: A meta-analy-sis, Ecology, 87(1), 53–63, doi:10.1890/04–1724.

Mackay, D. S., and L. E. Band (1997), Forest ecosystem processes at thewatershed scale: dynamic coupling of distributed hydrology and canopygrowth, Hydrol. Process., 11, 1197–1217.

Maggi, F., and A. Porporato (2007), Coupled moisture and microbial dy-namics in unsaturated soils, Water Resour. Res., 43, W07444,doi:10.1029/2006WR005367.

Maggi, F., C. Gu, W. J. Riley, G. M. Hornberger, R. T. Venterea, T. Xu, N.Spycher, C. Steefel, N. L. Miller, and C. M. Oldenburg (2008), A mecha-nistic treatment of the dominant soil nitrogen cycling processes: Modeldevelopment, testing, and application, J. Geophys. Res., 113, G02016,doi:10.1029/2007JG000578.


2570

http://www.agweather.geo.msu.edu/mawn/

http://www.agweather.geo.msu.edu/mawn/

http://daac.ornl.gov/SOILS/guides/igbp.html

http://eospso.gsfc.nasa.gov/atbd/modistables.html

http://eospso.gsfc.nasa.gov/atbd/modistables.html

Manzoni, S., and A. Porporato (2009), Soil carbon and nitrogen mineraliza-tion: Theory and models across scales, Soil Biol. Biochem., 41, 1355–1379, doi:10.1016/j.soilbio.2009.02.031.

Manzoni, S., and A. Porporato (2011), Common hydrologic and biogeo-chemical controls along the soil-stream continuum, Hydrol. Process., 25,1355–1360, doi:10.1002/hyp.7938.

Mao, D. Z., and K. A. Cherkauer (2009), Impacts of land-use change onhydrologic responses in the Great Lakes region, J. Hydrol., 374(1-2), 71–82, doi:10.1016/j.jhydrol.2009.06.016.

Mascaro, G., E. R. Vivoni, and R. Deidda (2010), Downscaling soil mois-ture in the southern Great Plains through a calibrated multifractal modelfor land surface modeling applications, Water Resour. Res., 46, W08546,doi:10.1029/2009WR008855.

Maxwell, R. M., and S. J. Kollet (2008a), Interdependence of groundwaterdynamics and land-energy feedbacks under climate change, Nat. Geosci.,1(10), 665–669, doi:10.1038/Ngeo315.

Maxwell, R. M., and S. J. Kollet (2008b), Quantifying the effects of three-dimensional subsurface heterogeneity on Hortonian runoff processesusing a coupled numerical, stochastic approach, Adv. Water Resour.,31(5), 807–817, doi:10.1016/j.advwatres.2008.01.020.

McDonnell, J. J., et al. (2007), Moving beyond heterogeneity and processcomplexity: A new vision for watershed hydrology, Water Resour. Res.,43, W07301, doi:10.1029/2006WR005467.

Michigan Department of Natural Resources, MNDR (2010), 2001 IFMAP/GAP lower Peninsula Land Cover. [Available at http://www.mcgi.state.-mi.us/mgdl/?rel¼thext&action¼thmname&cid¼5& cat¼LandþCoverþ2001, accessed 28 Nov. 2009.].

Miguez-Macho, G., and Y. Fan (2012), The role of groundwater in the Ama-zon water cycle: 2. Influence on seasonal soil moisture and evapotranspira-tion, J. Geophys. Res., 117, D15114, doi:10.1029/2012JD017540.

Miguez-Macho, G., Y. Fan, C. P. Weaver, R. Walko, and A. Robock(2007), Incorporating water table dynamics in climate modeling: 2. For-mulation, validation, and soil moisture simulation, J. Geophys. Res., 112,D13108, doi:10.1029/2006JD008112.

Mishra, V., K. A. Cherkauer, and S. Shukla (2010), Assessment of droughtdue to historic climate variability and projected future climate change inthe Midwestern United States, J. Hydrometeorol., 11(1), 46–68,doi:10.1175/2009jhm1156.1.

Mu, Q., F. A. Heinsch, M. Zhao, and S. W. Running (2007), Developmentof a global evapotranspiration algorithm based on MODIS and globalmeteorology data, Remote Sens. Environ., 111(4), 519–536,doi:10.1016/j.rse.2007.04.015.

Mu, Q., M. Zhao, and S. W. Running (2011), Improvements to a MODISglobal terrestrial evapotranspiration algorithm, Remote Sens. Environ.,115(8), 1781–1800, doi:10.1016/j.rse.2011.02.019.

NASA Land Processes Distributed Active Archive Center (LP DAAC)(2012), Land Processes Distributed Active Archive Center, MOD15A2,USGS/Earth Resources Observation and Science (EROS) Center, SiouxFalls, S. D., available at (https://lpdaac.usgs.gov/get_data)

National Climatic Data Center (NCDC) (2010), National Climatic DataCenter [Available at http://www.ncdc.noaa.gov/oa/climate/climatedata.html#daily.].

Neitsch, S. L., J. G. Arnold, J. R. Kiniry, and J. R. Williams (2005), Soiland Water Assessment Tool Theoretical Documentation Version 2005,U.S. Dep. of Agric.–Agric. Res Service, Temple, Tex.

Niu, G.-Y., and Z. L. Yang (2006), Effects of frozen soil on snowmelt runoff andsoil water storage at a continental scale, J. Hydrometeorol., 7(5), 937–952.

Niu, G.-Y., and Z.-L. Yang (2007), An observation-based formulation ofsnow cover fraction and its evaluation over large North American riverbasins, J. Geophys. Res., 112, D21101, doi:10.1029/2007JD008674.

Niu, G.-Y., C. Paniconi, P. A. Troch, R. L. Scott, M. Durcik, X. Zeng, T.Huxman, and D. C. Goodrich (2013), An integrated modelling frame-work of catchment-scale ecohydrological processes: 1. Model descrip-tion and tests over an energy-limited watershed, Ecohydrology, n/a-n/a,doi:10.1002/eco.1362.

Oishi, A. C., R. Oren, K. A. Novick, S. Palmroth, and G. G. Katul (2010),Interannual invariability of forest evapotranspiration and its consequenceto water flow downstream, Ecosystems, 13(3), 421–436, doi:10.1007/s10021-010-9328-3.

Oleson, K. W., et al. (2010), Technical Description of version 4.0 of theCommunity Land Model (CLM). NCAR Technical Note, Rep. NCAR/TN-478þSTR, National Center for Atmospheric Research, Boulder, Colo.

Oztan, M. (2011), GIS-enabled modeling of Michigan’s groundwater sys-tems, Ph.D. dissertation, Department of Civil & Environmental Engi-neering, Mich. State Univ., East Lansing.

Pan, Y.-X., and X.-P. Wang (2009), Factors controlling the spatial variabili-ty of surface soil moisture within revegetated-stabilized desert ecosys-tems of the Tengger Desert, Northern China, Hydrol. Process., 23(11),1591–1601, doi:10.1002/hyp.7287.

Pellenq, J., J. Kalma, G. Boulet, G.-M. Saulnier, S. Wooldridge, Y. Kerr,and A. Chehbouni (2003), A disaggregation scheme for soil moisturebased on topography and soil depth, J. Hydrol., 276, 112–127,doi:10.1016/S0022–1694(03)00066-0.

Qiu, Y., B. J. Fu, J. Wang, and L. D. Chen (2001), Soil moisture variationin relation to topography and land use in a hillslope catchment of theLoess Plateau, China, J. Hydrol., 240(3-4), 243–263, doi:10.1016/s0022–1694(00)00362-0.

Reich, P. B., S. E. Hobbie, T. Lee, D. S. Ellsworth, J. B. West, D. Tilman, J.M. H. Knops, S. Naeem, and J. Trost (2006), Nitrogen limitation con-strains sustainability of ecosystem response to CO2, Nature, 440(7086),922–925, doi:10.1038/nature04486.

Rihani, J. F., R. M. Maxwell, and F. K. Chow (2010), Coupling ground-water and land surface processes: Idealized simulations to identifyeffects of terrain and subsurface heterogeneity on land surface energyfluxes, Water Resour. Res., 46, W12523, doi:10.1029/2010WR009111.

Riveros-Iregui, D. A., B. L. McGlynn, R. E. Emanuel, and H. E. Epstein(2012), Complex terrain leads to bidirectional responses of soil respira-tion to inter-annual water availability, Global Change Biol., 18, 749–756, doi:10.1111/j.1365–2486.2011.02556.x.

Robertson, G. P., M. A. Huston, F. C. Evans, and J. M. Tiedje (1988), Spa-tial variability in a successional plant community: Patterns of nitrogenavailability, Ecology, 69(5), 1517–1524, doi:10.2307/1941649.

Sakaguchi, K., and X. Zeng (2009), Effects of soil wetness, plant litter, andunder-canopy atmospheric stability on ground evaporation in the Com-munity Land Model (CLM3.5), J. Geophys. Res., 114, D01107,doi:10.1029/2008JD010834.

Salvucci, G. D., and D. Entekhabi (1995), Hillslope and climatic controlson hydrologic fluxes, Water Resour. Res., 31(7), 1725–1739,doi:10.1029/95WR00057.

Schimel, D. S., B. H. Braswell, and W. J. Parton (1997), Equilibration ofthe terrestrial water, nitrogen, and carbon cycles, Proc. Natl. Acad. Sci.U. S. A., 94(16), 8280–8283, doi:10.1073/pnas.94.16.8280.

Sellers, P. J. (1985), Canopy reflectance, photosynthesis and transpiration,Int. J. Remote Sens., 6(8), 1335–1372.

Shen, C. P. (2009), A process-based distributed hydrologic model and itsapplication to a Michigan watershed, Ph.D. dissertation, 270 pp., Depart-ment of Civil & Environmental Engineering, Mich. State Univ., EastLansing.

Shen, C. P., and M. S. Phanikumar (SP10) (2010), A process-based, distrib-uted hydrologic model based on a large-scale method for surface-subsur-face coupling, Adv. Water Resour., 33(12), 1524–1541, doi:10.1016/j.advwatres.2010.09.002.

Simard, A. (2007), Predicting groundwater flow and transport using Michi-gan’s statewide wellogic database, Ph.D. dissertation, 109 pp., Depart-ment of Civil & Environmental Engineering, Mich. State Univ., EastLansing.

Sinha, T., and K. A. Cherkauer (2008), Time series analysis of soil freezeand thaw processes in Indiana, J. Hydrometeorol., 9(5), 936–950,doi:10.1175/2008jhm934.1.

Sivapalan, M. (2003), Prediction in ungauged basins: A grand challengefor theoretical hydrology, Hydrol. Process., 17(15), 3163–3170,doi:10.1002/hyp.5155.

Sivapalan, M., S. E. Thompson, C. J. Harman, N. B. Basu, and P. Kumar(2011), Water cycle dynamics in a changing environment: Improvingpredictability through synthesis, Water Resour. Res., 47, W00J01,doi:10.1029/2011WR011377.

Soil Survey Staff (2010), Soil Survey Geographic Database, NaturalResources Conservation Service, United States Dep. of Agric. [Availableat http://soils.usda.gov/survey/geography/ssurgo/, accessed 1 June2010.].

Sokolov, A. P., D. W. Kicklighter, J. M. Melillo, B. S. Felzer, C. A.Schlosser, and T. W. Cronin (2008), Consequences of considering car-bon-nitrogen interactions on the feedbacks between climate and the ter-restrial carbon cycle, J. Clim., 21(15), 3776–3796, doi:10.1175/2008jcli2038.1.

Suyker, A. E., and S. B. Verma (2009), Evapotranspiration of irrigated andrainfed maize-soybean cropping systems, Agric. For. Meteorol., 149(3-4), 443–452, doi:10.1016/j.agrformet.2008.09.010.

Tague, C. L., and L. E. Band (2004), RHESSys: Regional hydro-ecologicsimulation system—An Object-Oriented approach to spatially distributed


2571

http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=5&cat=Land+Cover+2001











https://lpdaac.usgs.gov/get_data

http://www.ncdc.noaa.gov/oa/climate/climatedata.html#daily

http://www.ncdc.noaa.gov/oa/climate/climatedata.html#daily

http://soils.usda.gov/survey/geography/ssurgo/

modeling of carbon, water, and nutrient cycling, Earth Interact., 8, 1–42,doi:10.1175/1087–3562(2004)8<1:RRHSSO>2.0.CO;2.

Thompson, S. E., et al. (2011), Patterns, puzzles and people: implementinghydrologic synthesis, Hydrol. Process., 25(20), 3256–3266, doi:10.1002/Hyp.8234.

Thornton, P. E., and N. A. Rosenbloom (2005), Ecosystem model spin-up:Estimating steady state conditions in a coupled terrestrial carbon andnitrogen cycle model, Ecol. Modell., 189(1-2), 25–48, doi:10.1016/j.ecolmodel.2005.04.008.

Thornton, P. E., and N. E. Zimmermann (2007), An improved canopy inte-gration scheme for a land surface model with prognostic canopy struc-ture, J. Clim., 20(15), 3902–3923, doi:10.1175/Jcli4222.1.

Toon, O. B., C. P. Mckay, T. P. Ackerman, and K. Santhanam (1989),Rapid calculation of radiative heating rates and photodissociation ratesin inhomogeneous multiple-scattering atmospheres, J. Geophys. Res.,94(D13), 16,287–16,301.

Troch, P. A., C. Paniconi, and E. E. van Loon (2003), Hillslope-stor-age Boussinesq model for subsurface flow and variable source areasalong complex hillslopes: 1. Formulation and characteristicresponse, Water Resour. Res., 39(11), 3-1–3-10, doi:10.1029/2002WR001728.

Turkoglu, N. (2010), Analysis of urban effects on soil temperature inAnkara, Environ. Monit. Assess., 169(1-4), 439–450, doi:10.1007/s10661-009-1187-z.

van Dam, J. C., and R. A. Feddes (2000), Numerical simulation of infiltra-tion, evaporation and shallow groundwater levels with the Richardsequation, J. Hydrol., 233(1-4), 72–85.

Vertessy, R. A., T. J. Hatton, R. G. Benyon, and W. R. Dawes (1996),Long-term growth and water balance predictions for a mountain ash (Eu-calyptus regnans) forest catchment subject to clear-felling and regenera-tion., Tree Physiol., 16, 221–232.

Vivoni, E. R. (2012), Spatial patterns, processes and predictions in ecohy-drology: Integrating technologies to meet the challenge, Ecohydrology,5(3), 235–241, doi:10.1002/eco.1248.

Vukovich, F. M. (1983), An analysis of the ground temperature and reflec-tivity pattern about St. Louis, Missouri, using HCMM satellite data,J. Clim. Appl. Meteorol., 22(4), 560–571.

Wagener, T., M. Sivapalan, P. A. Troch, B. L. McGlynn, C. J. Harman, H.V. Gupta, P. Kumar, P. S. C. Rao, N. B. Basu, and J. S. Wilson (2010),The future of hydrology: An evolving science for a changing world,Water Resour. Res., 46, W05301, doi:10.1029/2009WR008906.

White, M. A., P. E. Thornton, and S. W. Running (1997), A continentalphenology model for monitoring vegetation responses to interannual cli-matic variability, Global Biogeochem. Cycles, 11(2), 217–234.

Wilson, C. A., R. J. Mitchell, J. J. Hendricks, and L. R. Boring (1999), Pat-terns and controls of ecosystem function in longleaf pine-wiregrass sav-annas. II. Nitrogen dynamics, Canad. J. For. Res., 29(6), 752–760,doi:10.1139/cjfr-29-6-752.

Wilson, D. J., A. W. Western, and R. B. Grayson (2004), Identifying and quan-tifying sources of variability in temporal and spatial soil moisture observa-tions, Water Resour. Res., 40, W02507, doi:10.1029/2003WR002306.

Wutzler, T., and M. Reichstein (2008), Colimitation of decomposition bysubstrate and decomposers—A comparison of model formulations, Bio-geosciences, 5, 749–759.

Xu, R. and I. C. Prentice (2008), Terrestrial nitrogen cycle simulation witha dynamic global vegetation model, Global Change Biol., 14, 1745–1764, doi:10.1111/j.1365–2486.2008.01625.x.

Zeng, X., M. Zhao, and R. E. Dickinson (1998), Intercomparison of bulkaerodynamic algorithms for the computation of sea surface fluxes usingTOGA COARE and TAO Data, J. Clim., 11, 2628–2644, doi:10.1175/1520-0442(1998)011<2628:IOBAAF>2.0.CO;2.

Zeng, X., M. Shaikh, Y. Dai, R. E. Dickinson, and R. Myneni (2002), Cou-pling of the Common Land Model to the NCAR Community ClimateModel, J. Clim., 15, 1832–1854, doi:10.1175/1520-0442(2002)015<1832:COTCLM>2.0.CO;2.

Zhu, Q., and H. Lin (2011), Influences of soil, terrain, and crop growth onsoil moisture variation from transect to farm scales, Geoderma, 163(1-2),45–54, doi:10.1016/j.geoderma.2011.03.015.


2572

Evaluating controls on coupled hydrologic and vegetation dynamics ...

Documents

Transcript of Evaluating controls on coupled hydrologic and vegetation dynamics ...