CUAHSI Webinar, 13 November 2009 Water balance partitioning at the catchment scale: Random Process...

CUAHSI Webinar, 13 November 2009

Water balance partitioning at the catchment scale:

Random Process or Emerging Property?

Paul Brooks, Peter Troch, Ciaran Harman and Sally Thompson

CUAHSI Webinar13 November 2009



Motivation: another Horton index…

Horton, 1933 (AGU)

H constantV

W V : Growing-season vaporization (E+T)

W : Growing-season wetting (P-S)

“The natural vegetation of a region tends to develop to such an extent that it can utilize the largest possible proportion of the available soil moisture supplied by infiltration” (Horton, 1933, p.455)


Horton Index vs. Humidity IndexMean Horton Index Std. Horton Index

53% with Std(H)<0.0674% with Std(H)<0.0783% with Std(H)<0.0893% with Std(H)<0.10

Troch et al., 2009 (HP)


Objective: To address fundamental questions linking Hydrology and Ecology in a data-rich workshop setting

Hydrology

•Where does water go when it rains?

•What controls that partitioning?

Ecosystem Ecology

•How do we quantify plant available water?

•How does vegetation respond to changes in precipitation?

Can we improve hydrological, ecological, and biogeochemical predictability by introducing a reproducible measure of hydrologic partitioning into existing theory and observations?


Antoine Aubeneau, Ciaran Harman, Bryan Moravec, Andy Neal, Sally Thompson, Hal Voepel, Sheng Ye, Mary Yeager, Stefano Zanardo

A Selection of Results from the Summer Institute in Vancouver, BC


What controls the Horton index?


The Horton Index

Precip

“Fast” runoff

“Slow” runoff

ET

Wetting

Annual Evapotranspiration

Annual WettingHI =

Proportion of available water vaporized


Three approaches explain HI

FunctionProcessPattern

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Pre

dic

ted

HI_

50

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

HI_50

HI


... all three predict the mean remarkably well

ProcessFunctionPattern

Uncalibrated

Calibrated


HI was predictable based on static or mean catchment properties

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Pre

dic

ted

HI_

50

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

HI_50

Pattern

HI = f ( )

Humidity index P/EP

Mean Topographic Index<Log (a / tan β)>


Function

Functional model predicts mean, variance of HI

Wetting potentialFast flow threshold

P

S

U

ET

W

Functional model:

→ S and U have thresholds

→ ET and W have upper limit

…and using a conceptualization of annual partitioning of precip…


Process ... and using a stochastic model based on filtering of storm events.

Storagecapacity

Calibrated storage capacity

CalibratedUncalibrated


We gained insight into controls on HI


Empirical HI model

HUI

CT

I

0.1

0.2 0

.3

0.4

0.5

0.6

0.7

0

.8

0.9

1

1 2 3 4

34

56

78

Empirical CV(HI) model

HUI

CT

I

0

0.0

2

0.0

4

0.0

6

0.08

0.1

0.12

0.14

0.16

0.18

1 2 3 4

34

56

78

Regression models suggest that climate and topography are primary controlsPattern

Humidity IndexHumidity Index

Topographic Index

Mean: Climate (except in steep, arid regions)CV: topography (humid regions)

Mean HI CV HI


Functional model suggests catchment capacity to vaporize and store water are

basic controls

Ep λs = λu = 0

λs = λu = 0.05

Function

Mean: - vaporization potential (~ energy) - catchment “wetability” (to a point)

P = 1000mm


Process model also suggests keys are that climate and capacity to store water from storm eventsProcess

Mean HI: Humidity Index, storage capacityVariance: only sensitive in arid regions


Prediction of interannual variability opens up questions about other factors

Timing of rainfall, vegetation response, landscape change, …?

ProcessFunctionPattern


Key unresolved questions:

How does variability scale in time?

What timescales are important?


Key unresolved questions:

What is the role of vegetation in hydrologic partitioning?

Are we only able to make predictions because of the co-evolution of vegetation, soils and geomorphology constrained by climate, geology and time?


Variability and Vegetation

Learning from Data-Rich Sites


Working ParadigmClassic ecohydrological approach:

ETmax ~ f(Rn, VPD, LAI,T)

ET ~ ETmax * f(θ)

“Water-limited” paradigm? Plant control of ET?


A Parsimonious Model Penman Monteith

Model

Rn VPD LAI U P T

Emax

E

T

Interception Model

PPT

Runoff

Drainage

Infiltration

Multiple Wetting Front ModelRoot Water Uptake Model


Interannual variability


Sub-daily variability

ET

(m

m/h

r)


Seasonal variabilityE

T (

mm

/hr)

Month


Soil Moisture Drydown v ET

0 50 100 150 200 250 300 350 4000

0.5

1

1.5

2

2.5

3

E TS o il Mo is ture

800 900 1000 1100 1200 1300 1400-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Kendall

Sky OaksET increases as soil moisture declines! ET

Soil Moisture

ET correlates to soil moisture

Days

Days

ET

(m

m/h

r) o

r θ

%E

T (

mm

/hr)

or

θ %


Adding Groundwater Improves PredictionE

T (

mm

/hr)

Month


Phenology Changes Seasonality of ET

10 20 30 40 500

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

17

DOY

Nor

mal

ized

ET

, LA

I, R

n

L A I

E T

R n

0 50 100 1500

0.02

0.04

0.06

0.08

0.1

0.127

Radiation

ET

0 50 100 1500

0.02

0.04

0.06

0.08

0.1

0.129

Radiation

ET

0 1000

0.04

0.08

0.1213

Radiation

ET

A

B

C

A

B

C

Week

No

rmal

ized

ET

, LA

I an

d R

n

Howland Forest, Maine


Phenological Effects are Predictable

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.80

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

13

Norm alized Cum ulative GDD

Nor

mal

ized

ET

0 0.2 0.4 0.6 0.8 10.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

110


Nor

mal

ized

ET

0 0.1 0.2 0.3 0.4 0.5 0.6 0.70.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

111


Nor

mal

ized

ET

Kendall Grasslands Donaldson Coniferous Forest Morgan Monroe Mixed Forest

Poorly correlated Well correlated

ET v Cumulative Growing Degree Days for first 150 Days of the Year

Onset of plant growth?Or leaf maturity?


0 0.5 1 1.5 2 2.5 30

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

S ky O aks

Mo rg an Mo nro e

Harvard

G o o d win C re e k

Can Patches Predict Catchments?

Humidity Index

Ho

rto

n I

nd

ex

S.O. Catchment

M.M. Catchment

H.F. Catchment

G.C. Catchment

Sky Oaks

Morg. Monroe

Harvard Forest

Goodwin Crk.


Conceptual Upscaling Approach Multiple Buckets – different topography,

veg, soil etc.

PPT, Energy, C

ET, Energy, C

Deep Drainage, Water Table, Lateral Redistribution

Surface redistribution


Ecohydrological catchment classification?

Sky Oaks

Fort Peck Goodwin Creek

Howland Forest

Donaldson

Kennedy

Kendall

Austin Cary

Metolius

Harvard Forest

0.5 1 1.50

Morgan Monroe

Humidity Index

HuIRadiationPhenologyGW AccessSeasonality


Discussion Points

• What does all this mean for predicting water cycle dynamics in a changing environment?– Mean behavior of hydrologic partitioning is

surprisingly predictable, and– Knowing hydrologic partitioning improves

prediction of vegetation response, yet– The inter-annual variability is poorly understood

and calls for higher understanding of ecosystem control on water cycle dynamics (do we need to replace the old paradigm?)


Come see us at AGU!

Hydrologic Predictions in a Changing Environment

Monday 14th

Talks: 8:00 – 10:00 am and 10:20am – 12:20pm, 3005 Moscone West

Posters: 1:40 – 6:00pm Moscone South

CUAHSI Webinar, 13 November 2009 Water balance partitioning at the catchment scale: Random Process...

Documents

Transcript of CUAHSI Webinar, 13 November 2009 Water balance partitioning at the catchment scale: Random Process...