Comparison of satellite-derived potential evapotranspiration (MOD16A3) with in situ measurements from Quintana Roo, Mexico
Land surface evapotranspiration (ET) data is required accurately and in near-real time for operational applications such as regional drought monitoring. Satellitederived ET data products are promising for this purpose, particularly for poorly gauged areas such as south-eastern Mexico. Products (and algorithms) such as MODIS Global Evaporation Project (MOD16) are readily and publicly available but require evaluation prior to its regional and local application. Here, we present a comparison of MOD16 annual potential ET (PETMOD) with that computed from in situ measurements (34 weather stations) from eastern Yucatan Peninsula, for the years 2000 to 2012. A low correlation coefficient was found (r = 0.21) across sites and years. There was less temporal and spatial variation in PETMOD than ETo dataset. The range of ETo was: 453 - 2943 mm year-1, while that for PETMOD was: 1567 - 2353 mm year-1. The poor correlation found suggests that locally parameterized ET algorithms may provide better overall results than global MODIS data products.
| Main Authors: | , |
|---|---|
| Format: | Texto biblioteca |
| Language: | eng |
| Subjects: | Balance hídrico (Hidrología), Evapotranspiración, Artfrosur, |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Land surface evapotranspiration (ET) data is required accurately and in near-real time for operational applications such as regional drought monitoring. Satellitederived ET data products are promising for this purpose, particularly for poorly gauged areas such as south-eastern Mexico. Products (and algorithms) such as MODIS Global Evaporation Project (MOD16) are readily and publicly available but require evaluation prior to its regional and local application. Here, we present a comparison of MOD16 annual potential ET (PETMOD) with that computed from in situ measurements (34 weather stations) from eastern Yucatan Peninsula, for the years 2000 to 2012. A low correlation coefficient was found (r = 0.21) across sites and years. There was less temporal and spatial variation in PETMOD than ETo dataset. The range of ETo was: 453 - 2943 mm year-1, while that for PETMOD was: 1567 - 2353 mm year-1. The poor correlation found suggests that locally parameterized ET algorithms may provide better overall results than global MODIS data products. |
|---|