Keywords
Blue water, Green water, Tropical catchment
Start Date
17-9-2020 12:20 PM
End Date
17-9-2020 12:40 PM
Abstract
Water is becoming increasingly scarce resource worldwide, with agriculture being the most important consumer of water. Water consumption can be estimated by measuring evapotranspiration from remote sensing data. For water management purposes, it is essential to know how much of the evapotranspiration comes from rainfall (green water) or irrigation practices (blue water). In this study, we aim at evaluating several methods that aim at separating the blue and green water fluxes from evapotranspiration (ET) data for different Land Use Land Cover (LULC) classes of the Kikuletwa Catchment in Tanzania. Ensemble ET from seven global RS-based surface energy balance models: ETMonitor, GLEM, CMRS-ET, SSEBop, ALEXI, SEBS and MOD16 was used together with rainfall dataset CHIRPS as input to estimate blue and green water from total ET. We compared yearly blue water estimates from four different methods: ET-P (Senay et al., 2016), ET-Effective rainfall (Eekelen et al., 2015), the Budyko model and Soil Water Balance (SWB) method. The results show that all the methods have different spatial distribution of blue and green water. While three methods show forested areas to have highest blue water, SWB shows less blue water in the forested area but higher blue water in irrigated banana and coffee. Budyko method does not show a good agreement with other methods for all the LULC classes. The blue water estimates are higher than other methods. The data and methods presented in this study would provide guidance to water resource managers and planners of Pangani river basin in Tanzania supporting efficient planning, accounting and management of water use for agriculture taking into consideration that other LULC classes depend on the blue water.
Can we estimate Blue and Green water use for Different LULC of Tropical catchment in the Eastern Africa using Remote Sensing?
Water is becoming increasingly scarce resource worldwide, with agriculture being the most important consumer of water. Water consumption can be estimated by measuring evapotranspiration from remote sensing data. For water management purposes, it is essential to know how much of the evapotranspiration comes from rainfall (green water) or irrigation practices (blue water). In this study, we aim at evaluating several methods that aim at separating the blue and green water fluxes from evapotranspiration (ET) data for different Land Use Land Cover (LULC) classes of the Kikuletwa Catchment in Tanzania. Ensemble ET from seven global RS-based surface energy balance models: ETMonitor, GLEM, CMRS-ET, SSEBop, ALEXI, SEBS and MOD16 was used together with rainfall dataset CHIRPS as input to estimate blue and green water from total ET. We compared yearly blue water estimates from four different methods: ET-P (Senay et al., 2016), ET-Effective rainfall (Eekelen et al., 2015), the Budyko model and Soil Water Balance (SWB) method. The results show that all the methods have different spatial distribution of blue and green water. While three methods show forested areas to have highest blue water, SWB shows less blue water in the forested area but higher blue water in irrigated banana and coffee. Budyko method does not show a good agreement with other methods for all the LULC classes. The blue water estimates are higher than other methods. The data and methods presented in this study would provide guidance to water resource managers and planners of Pangani river basin in Tanzania supporting efficient planning, accounting and management of water use for agriculture taking into consideration that other LULC classes depend on the blue water.
Stream and Session
false