Accueil du site
Master
Pays Bas
2010
Estimation of actual evapotranspiration and water balance using combined geostationary and polar orbiting satellite products : a case study in Spain
Titre : Estimation of actual evapotranspiration and water balance using combined geostationary and polar orbiting satellite products : a case study in Spain
Auteur : Mamo, T.A. (Tadesse Alemu)
Etablissement de soutenance : University of Twente International Institute for Geo-Information Science and Earth Observation (ITC)
Grade : Master of Science in Geo-Information Science and Earth Observation 2010
Résumé
Actual evapotranspiration and rainfall are the key
elements in water balance estimation that give
scientifically sound information on water availability. Currently, satellite remote sensing is widely
used for estimation of these two parameters. In the
past decades, the quantification of daily actual
evapotranspiration (AET) from remote sensing data w
as mainly based on one time observation from
sun synchronous satellites by scaling instantaneous
evaporative fraction (EF) under the assumption of
constant daytime EF. Models and data source selections for accurate estimation of AET from remote
sensing observation are also continuously improving
. The sources of data for this thesis were 40
ground meteorological stations in the study area an
d satellite products from the European organization
for the exploitation of meteorological satellites on land surface analysis, satellite application facility
(EUMETSAT LSA SAF) and moderate resolution imaging
spectroradiometer (MODIS). Products
from EUMETSAT are mainly derived from the spinning
enhanced visible and infrared imager
(SEVIRI) radiometer embarked on the Meteosat second
generation (MSG) platform and other
European satellite systems which have an imaging-re
peat cycle of 15-30 minutes. For this research,
the rainfall products were taken from EUMETSAT meteorological product extraction facility-multi
sensor precipitation estimate (MPEF-MPE), but AET was estimated from remote sensing and ground-
based data using surface energy balance system mode
l (SEBS).
Efforts were made in this research to estimate daily average EF, instantaneous FE at different solar
times, EF from daily data, and daily AET for Castilla y Leon region in Spain using ground
meteorological stations, and combined high temporal
resolution geostationary satellite products and
relatively high spatial resolution polar orbiting satellite products. In addition, the performance of
SEBS model over different land cover classes was tested. The testing was only based on the
comparison of daily averaged EF and instantaneous E
F, and whether the results were reasonable or
not over all land cover classes in the study area.
The comparison of daily average EF with
instantaneous EF and EF from daily averaged data was done over different land cover classes for 15th
of March, 2009. In addition, SEBS estimated AET and
Meteosat second generation evapotranspiration
(MET) were compared with AET measured by an Eddy co
variance system in sparse vegetation. The
SEBS estimated AET was also compared with MET over
different land cover classes. Finally,
climatological water balance estimation “Precipitation minus evapotranspiration” was carried out over
rainfed cropland during the study period (March, 15
– September, 15, 2009) using rainfall products
from EUMETSAT and daily AET estimated by SEBS model
.
The results indicated that the average daily EF and
instantaneous EF have shown strong relation over
water bodies, irrigated croplands, rainfed croplands, shrubland, mosaic forest-shrubland and mosaic
crop-vegetation respectively, but it shows poor agreement over sparse vegetation, broadleaved
deciduous forest and needleleaved evergreen forest.
There was no agreement between daily average
EF and EF from daily averaged data in all land cover classes. It was observed that SEBS estimated
AET is close to the measured AET but MET overestimate. Only MET and SEBS estimated AET have
close values in irrigated croplands and water body,
but MET overestimated in the remaining land
cover classes. In addition, it was observed during
water balance estimation that there were only few
days in the months of May and June which have excess rainfall during the study period.
Page publiée le 27 janvier 2018