Mohsin Hafeez

Hydrological parameterization through remote sensing in Volta Basin, West Africa

Abstract Ground‐based hydrological data collection tends to be difficult and costly, especially in developing countries such as Ghana and Burkina Faso where the infrastructure for scientific monitoring is limited. Remote sensing has the potential to fill the gaps in observation networks. The GLOWA Volta Project (GVP) seeks to maximize the information to be gained from satellite imagery by combining remotely sensed data with strategically chosen ground observations. However, there is very limited information about the coupling of remotely sensed data with ground based data over the mixed savanna terrain of West Africa. This paper provides an overview of innovative techniques to measure hydrological parameters as actual evapotranspiration, rainfall, and surface runoff over mixed savanna terrain in a semi‐arid region in West Africa, and their potential use. Evapotranspiration – The Surface Energy Balance Algorithm for Land (SEBAL) was used to calculate sensible heat flux and evapotranspiration through the energy balance. The SEBAL parameterization is an iterative and feedback‐based numerical procedure that deduces the radiation, heat and evaporation fluxes. Along a 1,000 km gradient in the Volta Basin, three scintillometers were installed to measure sensible heat flux over distances comparable to NOAA‐AVHRR pixels, approximately two kilometers. The comparison of sensible heat flux measured from remotely sensed data and scintillometers provide accurate results. This will help to increase the reliability of SEBAL parameterization. Rainfall – Depending on the region within the Volta Basin, up to 90% of the precipitation in originates from squall‐lines. The Tropical Rainfall Measuring Mission (TRMM) imagery provides a valuable tool to monitor such squall lines. However, the TRMM signal should be validated for squall line rainfall. To increase the reliability of space‐based rainfall measurements, TRMM based rainfall rate estimates were calibrated with rainfall measurements from a dense network of rain gauges. Surface Runoff – Remote sensing has limited value in estimating surface runoff. The savanna of West Africa, however, is dotted with a large number of small reservoirs used to supply water for households, cattle, and small scale irrigation. Bathymetry of sixty reservoirs in Ghanas Upper‐East Region produced a very regular correlation between surface area, as observable by satellites, and volumes. By using all‐weather RADAR imagery and the measured surface/volume curves, surface runoff volumes can be monitored throughout the year. These indirect runoff measurements will help researchers to develop surface‐runoff models for the Volta Basin.

Modeling Irrigation Scheduling Under Shallow Groundwater Conditions as a Tool for an Integrated Management of Surface and Groundwater Resources

To restructure land- and irrigation-water use in Khorezm towards sustainability and economical feasibility, the current water use demands improvement. This requires increasing water use efficiency as much as possible, while at the same time minimizing negative impacts on the production system. These objectives can be reached with an integrated management of the irrigation and drainage system. To develop optimal management strategies, models describing the water distribution (irrigation scheduling model) and analyzing the impact on the groundwater (groundwater models) will be very helpful. In the Water Users Association (WUA) Shomakhulum, located in the southwest of Khorezm and with an irrigated area of approximately 2,000 ha, current irrigation strategies were monitored. Overall irrigation efficiency of the sub-unit representing the WUA is rather low (33%). Besides the poor state of the irrigation infrastructure, major reasons for the low efficiency are on the one hand a lack of detailed and up-to-date information on the system and its temporal behavior, and on the other hand missing options to consider detailed information in the procedures to establish water distribution plans. To tackle these issues, the irrigation scheduling model FAO CROPWAT was applied as an alternative to the current rather rigid water distribution planning. Feeding the model with detailed information on the irrigation system and its behavior (application efficiency by field-water balancing, network efficiency based on ponding experiments) provided a powerful tool to improve water use. As the groundwater in Shomakhulum is shallow, the model was further developed in order to assess the importance of the capillary rise. Therefore, the soil-water model Hydrus-1D was applied. The results of the study show that capillary rise is an important factor in water balancing and can contribute a maximum of 28% of crop-specific evapotranspiration in cotton, 12% in vegetables and 9% in winter wheat. In-practice irrigation scheduling, when simulated and assessed with the CROPWAT model, showed a 7–42% reduction in cotton yield. If the overall irrigation efficiency is improved to 56%, water saving of 41% can be achieved. Introducing alternative crops to cotton can result in 6% water saving. About 15–20% of the water can be saved by leaving marginal lands, i.e., land of low quality, out of the production.

Comprision study on mapping of ET in the CIA of Murrumbidgee catchment with remote sensed satellite data: Examples from National Airborne Field Experimentation

This study focused on actual Evapotranspiration (ET) characterization in Coleambally Irrigation Area (CIA), Murrubidgee Irrigation Area (MIA) and its ambient region using the SEBAL model with remotely sensed TERRA/MODIS, NOAA/AVHRR and Landsat-5 TM data in the Area Of Interest (AOI) for the National Airborne Field Experimentation (NAFE06) campaign in 2006. Results showed that actual ET estimated from NOAA AVHRR 18 was always overestimating as comparison to Eddy system (on average 38%) during the image acquisition dates. However, for the same image acquisition dates, TERRA/MODIS ET ranges from 3.7% lower to 21.7% higher than the Eddy system. Landsat 5 TM modeled ET results were comparable to the Eddy Covariance system having a minor error with an average of 4.09%. It was proven possible to simultaneously use SEBAL for different sensors with the combination of high spatial and temporal resolution to estimate ET spatial distribution characteristics though the accuracy of NOAA-AVHRR derived result is not ideal. Considering the lack of high spatial resolution thermal satellite and need of time-series ET dynamics, the MODIS data could be used to provide seasonal actual ET for regional studies. The combination of MODIS and Landsat can be a better choice for future ET study at catchments scale.