 |
| Titel |
Use of digital images to observe forest phenology and drought stress |
| VerfasserIn |
H. E. Ahrends, S. Etzold, W. Eugster, N. Buchmann, F. Jeanneret, H. Wanner |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250028832
|
|
|
|
|
|
| Zusammenfassung |
| Phenological data that complement research studies of climate impacts on ecosystems need to
be estimated with both temporal and spatial accuracy. Forest phenology can be monitored by
satellite, but the realism of remote sensing products such as the NDVI (Normalized
Difference Vegetation Index) still heavily depends on ground based validation data. Ground
based data is often observer-biased and the number of observations strongly varies in time
and space. Recent studies have demonstrated the successful application of digital
camera images for spring phenological monitoring in ecosystem studies. Objective
of the present study therefore was to test the application of digital images from
standard RGB-cameras for regional monitoring and modelling the seasonality of forest
physiology and for detecting species-specific reactions on environmental impacts such as
drought.
A digital camera was mounted on the uppermost platform of a fluxtower at the
CarboEurope site Lägeren (northern Switzerland). Daily images of the mixed forest from
four years were used to derive the timing of greenup, leaf maturity, senescence and
dormancy of two different tree species (beech and ash) between 2005 and 2008.
Based on the image color values a vegetation index was computed. Time series of
the vegetation index were jointly analyzed with standard meteorological data and
eddy covariance measurements of ecosystem carbon dioxide and water vapour
exchange.
Generally the observation of phenologial phases was successful but complex for the end
of the vegetation period, e.g. due to early leaf coloring caused by summer heat, and a less
pronounced starting date of leaf senescence compared with spring greenup. Spring CO2 flux
characteristics could be explained by leaf emergence dates of dominant tree species. A
drought period in 2006 influenced index values for beech but not for the highly
drought-tolerant ash trees. Phenological data showed significant correlation with
carbon dioxide exchange measurements and the computed index largely explained
the seasonal and interannual variability of the forests gross primary productivity.
We conclude good applicability of digital cameras for a) observing phenological
phases and b) complementing data to model ecophysiological processes of a forest. |
|
|
|
|
|
|