Phenological monitoring across the Northern Forest using a network of digital webcams

Theme #2:Principal Investigator(s): Andrew Richardson

Affiliations/Institutions: UNH/HarvardEmail: arichardson@oeb.harvard.edu

Mailing address: Department of Organismic and Evolutionary Biology, Harvard University, 22 Divinity Ave, Cambridge MA

Collaborators and affiliations: too many to list here!Completion date: December 2009

Funding support for this project was provided by the Northeastern States Research Cooperative (NSRC), a partnership of Northern Forest states (New Hampshire, Vermont,

Maine, and New York), in coordination with the USDA Forest Service.http://www.nsrcforest.org

We deployed webcams to a dozen research sites across the northeastern US and adjacent Canada, and we are tracking seasonal cycles of canopy development and senescence by conducting quantitative analysis of image “greenness”.

Collaborators and CooperatorsDavid Y. Hollinger, USDA Forest Service, Durham, NH (davidh@hypatia.unh.edu)John Campbell, USDA Forest Service, Durham, NH (jcampbell@lternet.edu)Peter Groffman, Institute of Ecosystem Studies, Millbrook, NY (groffmanp@ecostudies.org)Hank Margolis, Laval University, Sainte-Foy, Quebec (hank.margolis@sbf.ulaval.ca)William Munger, Harvard University, Cambridge, MA (jwm@io.harvard.edu)John O’Keeefe, Harvard Forest, Petersham, MA (jokeefe@fas.harvard.edu)Timothy Perkins, University of Vermont, Burlington, VT (timothy.perkins@uvm.edu)Steven Wofsy, Harvard University, Cambridge, MA (swofsy@deas.harvard.edu)Amey Bailey, USDA Forest Service, Campton, NH (ameybailey@fs.fed.us)Neal Scott, Queen’s University, Kingston, Ontario (scottn@post.queensu.ca)Myron Mitchell, SUNY-ESF, Syracuse, NY (mitchell@mailbox.syr.edu)Ellen Denny, USDA Forest Service, Durham, NH (ellen.denny@yale.edu)Brenden McNeil, University of Wisconsin, Madison, WI (bmcneil@wisc.edu)

Project SummaryPhenology is the study of recurring lifecycle events influenced by seasonal environmental

changes. For deciduous forests, critical phenological events are spring development and autumn senescence of the canopy, as these effectively mark the beginning and end, respectively, of the growing season. Phenology thus exerts a strong control on both spatial and temporal patterns of primary productivity and is therefore critical to carbon sequestration of terrestrial ecosystems. Beyond carbon, however, phenology also influences hydrologic processes, as spring leaf-out is accompanied by a marked increase in evapotranspiration, and nutrient cycling processes as autumn senescence results in fresh litter (nutrient) inputs to the forest floor. Phenology has been shown to be a robust integrator of the effects of climate change (e.g., recent warming trends) on natural systems, and it is now recognized that improved monitoring of phenology on local-to-continental scales is needed. These data will contribute to reducing uncertainties about the effects of climate change and variability on ecosystem processes and forest health.

We initiated a research network that would provide automated “near surface” remote sensing of canopy phenology across the Northern Forest region. We installed commercial-grade digital webcams at 10 established research sites within this area, from Ontario and New York across to Maine at a latitude of roughly 45°N. We are using images from these cameras to track both spring green-up and autumn senescence based on changes in red, green, and blue (RGB) color channel brightnesses of each image. At five of the proposed research sites, ongoing measurements of carbon and water fluxes are being made with the eddy covariance method, and we are directly linking phenology to ecosystem processes. Images from the camera network are presented at the PhenoCam project web page, http://phenocam.sr.unh.edu.

This cooperative effort involved researchers active in five states and two Canadian provinces, and thus contributes to the NERC objective of cross-border collaboration. Our project will provide valuable information to forest managers (monitoring of forest health) and the tourism industry (targeting optimal “leaf peeping” conditions).

Background and JustificationPhenology has been shown to be a robust integrator of the effects of year-to-year climate variability and longer-term climate change on natural systems (e.g., recent warming trends). There is a need to better document biological responses to a changing world, and improved phenological monitoring at scales from individual organisms to ecosystems, regions and continents will contribute to achieving this goal. The USA-National Phenology Network is being designed and organized to engage federal agencies, environmental networks and field stations, educational institutions, and mass participation by citizen scientists in the development of a continental-scale phenological monitoring program. The PhenoCam network is actively working with USA-NPN and participating in this effort.

We initiated PhenoCam in order to provide automated, near-surface remote sensing of canopy phenology across the Northern Forest region of New England, upstate New York, and adjacent Canada. Image analysis of archived webcam images provides an objective means by which canopy phenology can be monitored and quantified, at relatively low cost and with minimal personnel expenses, without the need for a human observer. The monitoring network will, over time, provide a unique and scientifically valuable long-term data set on seasonal patterns of phenology across the Northern Forest.

Methods• Webcam phenology: a form of

“Near surface” remote sensing• Commercially available webcam

mounted on tower– Faces north– 15° below horizontal– Spatial integration (but individual crowns

could be analyzed)– Images recorded every 30 minutes,

uploaded by ftp to web page• Provides a permanent visual record • Image analysis (RGB channel extraction)

to quantify phenological changes in “greenness” and “redness”

• Not a calibrated instrument—but neither are field observers!

Results/Project outcomes

Vertical lines in graphs correspond to images shown above (Bartlett Experimental Forest)

Quantitative analysis of images: relative brightness of red (%Red) and green (%Green) channels show strong seasonal patterns:

• Timing and speed of spring green-up and autumn green-down

• Timing and peak intensity of autumn coloration

With data from the PhenoCam network, we are looking at these patterns in a regional context, across sites spanning 10°C in mean annual temperature. Sp

ring

rise

Aut

umn

sene

scen

cePe

ak c

olor

s

Camera greenness in relation to phenological changes on the ground

Figure shows seasonal trajectory of canopy greenness derived from Harvard Forest webcam pictures. A clear rising trend begins in spring, around day 120, coinciding with budburst; greenness peaks as the canopy nears complete development by day 150. A relatively flat summertime plateau of stable greenness persists until day 240, at which time the canopy begins to yellow with autumn senescence. Greenness declines over the following 60 days, as autumn reds and browns come to dominate the canopy, prior to complete abscission. Vertical lines indicate observed dates (courtesy J O’Keefe) of budburst, 75% leaf development, and 50% leaf fall for dominant species.

Implications and applicationsin the Northern Forest region

• Our project improves regional-scale understanding of inter-relationships among phenology, canopy state, ecosystem function (particularly C cycling), and abiotic factors. This knowledge is essential for predicting the effects of climate change on both ecosystem productivity and forest health, a key objective of NSRC Theme 2

• Results from our project are directly relevant to a number of industries in the Northern Forest region, including forest products (land managers want to know how climate change will affect stand productivity and C sequestration) and tourism (with PhenoCam, quantitative, real-time tracking of peak colors can be used to target where “leaf peepers” should go for the best viewing experience).

• This cooperative effort involved researchers active in five states and two Canadian provinces, and thus contributes to the NERC objective of cross-border collaboration.

Future directions

• We are conducting a number of synthesis projects that draw on our ever-expanding PhenoCam image archive:

• Focus #1: Analyzing relationships between phenology and spatio-temporal variation in forest productivity: at two-thirds of the PhenoCam sites, cooperating researchers are measuring surface-atmosphere CO2exchange using the eddy covariance method, and these flux data will be used to study the implications of seasonal changes in canopy state on C cycling and sequestration

• Focus #2: Link the seasonal changes in webcam greenness to the long-term observer records of phenological events (e.g., budburst and abscission) being maintained at many sites (Harvard Forest, Hubbard Brook, PMRC, etc.)

• Focus #3: Use phenological transition dates extracted from the webcam data as ground validation for satellite remote sensing phenology products (e.g. from MODIS).

List of products #1Peer reviewed publicationsRichardson, A.D., B.H. Braswell, D.Y. Hollinger, J.P. Jenkins and S.V. Ollinger. 2009. Near-surface remote

sensing of spatial and temporal variation in canopy phenology. Ecological Applications, 19: 1417-1428.

Jacobs, N., W. Burgin, N. Fridrich, A. Abrams, K. Miskell, R. Braswell, A. Richardson and R. Pless. 2009. The global network of outdoor webcams: properties and applications. Proceedings ACM GIS '09 (November 4-6, 2009. Seattle, WA, USA), 111-120.

White, M.A., K.M. de Beurs, K. Didan, D.W. Inouye, A.D. Richardson, O.P. Jensen, J. O’Keefe, G. Zhang, R.R. Nemani, W.J.D. van Leeuwen, J.F. Brown, A. de Wit, M. Schaepman, X. Lin, M. Dettinger, A.S. Bailey, J. Kimball, M.D. Schwartz, D.D. Baldocchi, J.T. Lee, W.K. Lauenroth. 2009. Intercomparison, interpretation, and assessment of spring phenology in North America estimated from remote sensing for 1982 to 2006. Global Change Biology, 15: 2335-2359.

Morisette, J.T., A.D. Richardson, A.K. Knapp, J.I. Fisher, E.A. Graham, J. Abatzoglou, B.E. Wilson, D.D. Breshears, G.M. Henebry, J.M. Hanes, and L. Liang. 2009. Tracking the rhythm of the seasons in the face of global change: phenological research in the 21st Century. Frontiers in Ecology and the Environment, 7: 253-260.

Richardson, A.D. and J. O’Keefe. 2009. Phenological differences between understory and overstory: A case study using the long-term Harvard Forest records. In: A. Noormets, ed. Phenology of Ecosystem Processes. Springer Science + Business, New York. pp. 87-117.

List of products #2Other publicationsHenebry, G.M., A.D. Richardson, D.D. Breshears, J. Abatzoglou, J.I. Fisher, E.A. Graham, J.M. Hanes, B.E.

Wilson and J.T. Morisette. 2009. Distinguishing among factors driving phenologies: a reply to Sagarin. Frontiers in Ecology and the Environment, 7:296. doi:10.1890/09.WB.021.

Wingate, L., A.D. Richardson, J.F. Weltzin, K.N. Nasahara, J. Grace. 2008. Keeping an eye on the carbon balance: linking canopy development and net ecosystem exchange using a webcam. FluxLetter (The Newsletter of FLUXNET) 1(2): 14-17.

McNeil, B., E. Denny, and A.D. Richardson. 2008. Coordinating a Northeast Regional Phenology Network. Bulletin of the Ecological Society of America 89: 188-190.

Web pages

PHENOCAM, http://phenocam.sr.unh.edu/

Podcast

Ecological Society of America, Beyond the Frontier: “In Ecology, Timing is Everything” (June 9, 2009); available at http://www.esa.org/podcast/.

List of products #3Conference Presentations, Posters and Abstracts:Richardson, A.D., and FLUXNET Collaborators. 2009. The effects of the timing of spring and autumn phenological switches on forest-

atmosphere carbon exchange. Ecological Society of America, Annual Meeting, Albuquerque NM (abstract, oral presentation)Wingate, L., A.D. Richardson, K.N. Nasahara, J.F. Weltzin, J. Grace and the FLUXNET Webcam Team. 2009. Keeping an eye on the

carbon balance: linking canopy development and net ecosystem exchange using an international webcam network. European Geosciences Union, General Assembly, Vienna, Austria (abstract, oral presentation)

Braswell, B.H., A.D. Richardson, S.V. Ollinger, M.A. Friedl, and D.Y. Hollinger. 2009. Combining observations of a digital camera network, satellite remote sensing, and micrometeorology for improved understanding of forest phenology. European Geosciences Union, General Assembly, Vienna, Austria (abstract, oral presentation)

Richardson, A.D., B.H. Braswell, J.P. Jenkins, and D.Y. Hollinger. 2009. Monitoring canopy phenology at a regional scale with a network of digital webcams. COST 725 Action Meeting on Phenology, Geisenheim, Germany (abstract, oral presentation).

Richardson, A.D., J.P. Jenkins, B.H. Braswell, D.Y. Hollinger and S.V. Ollinger. 2009. Detecting the variability in canopy phenology with a network of digital cameras. 2nd NACP All-Investigators Meeting, San Diego CA (abstract, poster).

Richardson, A.D., B.H. Braswell, D.Y. Hollinger, J.P. Jenkins and S.V. Ollinger. 2008. Near surface remote sensing of spatial and temporal variation in canopy phenology. AmeriFlux Annual Meeting, Boulder CO (poster, abstract). [Also presented as a poster at the 2008 Northeastern Ecosystem Research Cooperative Conference, Durham NH.]

Benton, L.M., J.F. Weltzin, A.D. Richardson, and M.V. Losleben. 2008. Towards a nation-wide phenology camera monitoring network.Ecological Society of America Annual Meeting, Milwaukee WI (poster, abstract).

Richardson, A.D., D.Y. Hollinger, B.H. Braswell and J.P. Jenkins. 2008. Spring comes to Bartlett/And autumn senescence too/Omnisicent webcam: Monitoring regional patterns of canopy-scale phenology with a network of digital webcams. Fluxnet-Canada/Canadian Carbon Program Annual Meeting, Calgary AB (invited oral presentation).

Richardson, A.D., B.H. Braswell, and D.Y. Hollinger. 2008. Phenological monitoring across the Northern forest region using a network of digital webcams. 19th Annual Harvard Forest Ecology Symposium, Petersham MA (abstract).

Verma, M., M. Friedl, N. Phillips, and A.D. Richardson. 2008. Comparing field measurements and vegetation indices in support of moderate resolution remotely sensed phenology. International Geoscience and Remote Sensing Symposium (IGARSS), Boston MA (abstract).

Richardson, A.D., B.H. Braswell, D.Y. Hollinger, and J.P. Jenkins. 2007. Monitoring regional patterns of canopy-scale phenology with a network of digital webcams. AGU Fall Meeting, San Francisco CA (abstract, invited oral presentation).

Richardson, A.D. 2007. Linking interannual variation in carbon fluxes to phenology in forests of the northeastern United States.AmeriFlux Annual Meeting, Boulder CO (invited oral presentation).

List of products #4Departmental Seminars

Max-Planck-Institut für Biogeochemie, Jena, Germany, June 2009.Technische Universität München, Forstwissenschaften und Ressourcenmanagement, Fachgebiet für Ökoklimatologie, Freising,

Germany, May 2009.Massachusetts Institute of Technology, Lincoln Laboratory, Lincoln MA, March 2009.Marine Biological Laboratory, Ecosystems Center, Woods Hole MA, January 2009.Boston University, Department of Geography and Environment, Boston MA, November 2008.Harvard University, Department of Organismic and Evolutionary Biology, Cambridge MA, March 2008.Edinburgh University, School of GeoSciences, Edinburgh, Scotland, September 2007.

Press coverage

UNH Alumni Magazine, Poughkeepsie Journal, Canadian Carbon Program Newsletter, Minneapolis Star-Tribune, Kingston Whig-Standard

TV Interview

Mish Michaels, WBZ-TV Boston