Evaluation of the Seasonal Variability of Evapotranspiration using Micro- Meteorological Observations and Remote Sensing Techniques in Ghana and Burkina-Faso.

Abstract

The availability and management of water supply in the sub-Sahara region of West-Africa is becoming a great concern. Thus, adequate information on the spatial and temporal distribution of surface energy balance components in these areas are critical information for sustainable management of water resources as well as for a better understanding of water and heat exchange processes between the land surface and the atmosphere. However the distributions of surface fluxes over large watershed cannot be obtained from ground measurements alone. Therefore their prediction from remote-sensing data is essential as it can provide information of the energy and water balance components on a regional scale. The specific objectives of this study were to; determine the sensible and latent heat fluxes in Vea and Tono watersheds (Ghana) and Nazinga park in Burkina-Faso and estimate evapotranspiration and energy fluxes within the catchments. The methodology involved the use of EC methods and remote sensing algorithm in estimating the surface energy fluxes and the daily evapotranspiration. Three Eddy Covariance stations were installed close to the Ghana–Burkina-Faso border in Sumbrungu Aguusi, (10.841°N, 0.918°W), Kayoro Dakorenia (10.918°N, 1.319°W), both in Ghana, and the third station was installed in the Nazinga Park (11.152°N, 1.586°W), Burkina-Faso, West Africa. Data from the three stations were used to determine the surface energy fluxes. Also, a remote sense technique SEBAL (Surface Energy Balance Algorithm for Land) model was selected to estimate the regional distribution of the energy balance components and the evapotranspiration from these sites. Comparison of the remote sensing/SEBAL derived evapotranspiration with actual Eddy Covariance measurements was done. Results showed that the Incoming shortwave radiation, out-going shortwave radiation, incoming long wave radiation and out-going longwave radiation changed seasonally, resulting in a seasonal variation in net radiation. It was observed that the soil moisture content and evaporative fraction was low during the dry season but high during the wet season, whereas, the surface albedo and the Bowen ratio was low during the dry season but high during the wet season. Most of the available energy were converted to latent heat during the wet season, while, during the dry season, most of the available energy were converted to sensible heat. The closure of the energy balance was examined at the three stations; slopes of the regression obtained were 0.67, 0.66 and 0.87, with intercept of 33 Wm-2, 11 Wm-2 and 10.71 Wm-2 for the three stations respectively. The causes of the non-closure of the stations energy balance were attributed to; instrumentation of the individual energy component measurements; the footprint from the individual energy components are not consistent; and the transport of large scale eddies which cannot be measured with an eddy covariance system. Using the SEBAL model to estimate surface fluxes and daily evapotranspiration, the results revealed that the energy fluxes and evapotranspiration varied in accordance to different land use and land cover. When compared with those obtained from the EC system, which were point measurement, the SEBAL results showed relatively good correlation of r2 (coefficient of determination) of 0.99, 0.97 and 0.99 the instantaneous net radiation, r2 of 0.62, 0.61 and 0.83 for the sensible heat flux, r2 of 0.83, 0.76 and 0.68 for the latent heat, r2 of 0.83, 0.77 and 0.77 for the daily evapotranspiration at the Sumbrungu, Kayoro and Nazinga park respectively. However, r2 of 0.03, 0.03 and 0.17, was observed between the modelled and measured ground heat flux, this is due to the fact that the point measured ground heat flux may not be able to represent the size of a satellite pixel. The study therefore, concluded that the SEBAL algorithm has the potential of providing spatial surface energy fluxes and daily ET on clear-sky days over large areas. In a situation where there are insufficient in situ observations and less meteorological stations network that can give better description of the climate variability of a region and validate climate model predictions, evapotranspiration modelled from remote sensing data using the SEBAL algorithm could be an alternative in providing decision support for system-wide operational water management issues.