|
Titel |
Using dual temperature difference two source energy balance model and MODIS data to estimate surface energy fluxes at regional scales in northern latitudes |
VerfasserIn |
R. Guzinski, M. Anderson, W. Kustas, H. Nieto, I. Sandholt |
Konferenz |
EGU General Assembly 2012
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 14 (2012) |
Datensatznummer |
250065244
|
|
|
|
Zusammenfassung |
A Two Source Energy Balance (TSEB) thermal-based modeling scheme has previously been
used to successfully estimate surface latent and sensible heat fluxes at regional to continental
scales with the help of satellite surface radiometric temperature observations. The Dual
Temperature Difference (DTD) model introduced a simple methodology to address the
sensitivity of the thermal-based energy balance models to the absolute measurement of land
surface temperature (LST), which when derived with the help of satellites can have errors of
several degrees. The original DTD model formulation required an early morning LST
observation (1 hour after local sunrise) when fluxes were minimal followed by another LST
observations later in the morning or afternoon and so was limited in use to data provided by
geostationary satellites having high temporal resolution. This, however, made it
unsuitable for areas at higher latitudes, such as northern Eurasia and northern North
America.
In this poster we present a number of modifications to the DTD model which allows it to
exploit the day and night LST observations by the MODIS sensor aboard the Terra and Aqua
polar orbiting satellites. Firstly, we look at whether taking the first LST observation around
the time of Aqua’s night overpass, when fluxes are small but not insignificant, would greatly
affect the accuracy of the model. Secondly, we consider the issues directly related to using the
MODIS sensor to measure the LST. This includes different view zenith angles of the day
and night LST observations, the two observations possibly coming from the two
different satellites and the accuracy of the instrument itself. We also evaluate two
approaches for estimating αPT, the Priestley-Taylor parameter used in the TSEB
modeling scheme to estimate heat fluxes of the vegetation canopy, to improve the
performance of the model in coniferous and deciduous forests. The first approach
estimates αPT based on tree height, while the second uses fraction of vegetation
that is green, calculated from vegetation indices as a multiplier of a default αPT
value.
We run the modified DTD model with LST observations taken by the MODIS instrument
and validate the output against field measurements in a number of different ecosystems in
Denmark and USA obtaining good accuracy of the modeled sensible and latent heat fluxes.
Finally we produce regional maps of energy fluxes over the area of a hydrological
observatory in western Denmark, HOBE. |
|
|
|
|
|