Analysis of monthly MOD16 evapotranspiration rates at sites with different climatic characteristics; Heuningnes and Letaba catchments in South Africa
Loading...
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Western Cape
Abstract
Evapotranspiration (ET) is an important component of the water cycle that is estimated to return
about 60% of precipitation back to the atmosphere. Actual ET can be estimated using remote
sensing techniques and ground-based measurements. In recent years, a remote sensing product
MOD 16 ET has been developed. The limited validation of this product done in South Africa
showed that ET was underestimated at some sites. A comprehensive analysis of historic and
seasonal trends in MOD 16 ET data in different climatic regions of South Africa has not been
done.
This study has the objective of evaluating if MOD 16 evapotranspiration estimates realistically
represent the seasonal variations of ET on different land cover types in two different climatic
regions; Mediterranean (Heuningnes catchment in Western Cape) and Sub-tropical (Letaba
catchment in Limpopo) regions. Monthly MOD 16 ET maps for 2000-2012 for the Letaba
catchment and Heuningnes catchment were created using ArcGIS. The results suggested that
during the 2000 - 2012 period, ET was 438 - 753 mm/ year in the Letaba catchment and 432 -
458 mm/year in the Heuningnes catchment.
The accuracy of MOD 16 ET was evaluated using estimates of actual ET from scintillometer
data in Elandsberg (Western Cape) and flux tower data in Malopeni (Limpopo) and Skukuza
(Limpopo). Monthly ET estimated using scintillometer and flux tower were calculated to
coincide with the monthly MOD 16 ET data for a period of 1 year from Nov 2012- Oct 2013 in
Elandsberg, 1 year 1 month from Mar 2009- Mar 2010 in Malopeni and 13 years from 2000-
2012 in Skukuza. In Elandsberg, the results showed that MOD 16 underestimated ET (R2 = 0.16,
RMSE = 28.30 mm/month). In Malopeni, the results suggested that there is a strong relationship
between ET estimated from flux tower data and MOD 16 ET (R2 = 0.77), but MOD 16 slightly
overestimate ET (RMSE = 8.6 mm/month). MODIS ET estimates for Elandsberg had a poorer
comparison with the results obtained at Malopeni and Skukuza.
Description
Magister Scientiae - MSc (Environ & Water Science)