Use Of Remote Sensing To Estimate Evapotranspiration (Et) For Water Resources Management : Case Study Ndaka-Ini Dam Watershed

Abstract

Water is a very important commodity for sustaining life and also a key factor in sustaining agricultural production, energy production and other activities at optimal levels. Despite water being very important, the world including Kenya is far from being water secure which is proven by the demand of water already surpassing supply in many regions. Water availability has been acknowledged as a global problem and requires to be consistently assessed to support sustainable use. Assessing water availability is one aspect of water resources management. Evapotranspiration (ET) is the most problematic constituent of the water cycle to estimate precisely due to the heterogeneity of the landscape and the big number of controlling factors. Generally, the methods of obtaining ET are classified into three groups including direct measurement, modelling methods and Remote sensing methods. The overall objective of this study was to estimate spatial and temporal evapotranspiration for water resource assessment and management using remote sensing. Specifically, it sought to identify parameters for estimating evapotranspiration, select suitable remote sensing data for estimating evapotranspiration and finally estimate spatial and temporal evapotranspiration in Ndaka-ini Dam watershed between the years 2016-2019.The project utilized Surface Energy Balance system Model that requires emissivity, Land Surface Temperature, Albedo, Fractional Vegetation cover, Leaf Area Index (LAI), and Normalized Vegetation Difference Index (NDVI) inputs obtained from remote sensed images. Landsat 8 images of 30 metres resolution were found adequate for the study. The Maps showing spatial and temporal estimated values per Land-cover in Ndaka-ini watershed between the years 2016-2019 were obtained. The results showed that the Mean estimated ET values were 5.65, 4.67, 6.23 and 6.07mm/day for 2016, 2017, 2018 and 2019 respectively. These indicated that 2018 had the highest mean estimated ET which is an indication of more water availability in the catchment as compared to the other years. Spatio-temporal variations in ET for the different land cover results indicate that dense forest has the lowest mean ET of 5.39mm/day and Perennial cropland has the highest mean ET of 5.83 mm/day. This is a clear indication that some land covers like perennial cropland has greater water requirements that others. It was concluded the spatial and temporal distribution of estimated ET were mapped which is an advantage of this method over the direct measurements method. Remote sensing could hence be used to estimate ET at regional and global scales which can help in water accounting and planning.