MODELING CONCEPTS AND REMOTE SENSING METHODS FOR SUSTAINABLE WATER AND LAND MANAGEMENT OF THE OKAVANGO DELTA, BOTSWANA
INTRODUCTION & RESEARCH PROBLEM:
The Okavango Delta (~30,000 km2) is one of the world’s largest inland wetland ecosystems, situated in the northwestern
Botswana, Southern Africa. The semi-arid countries of Angola, Namibia and Botswana share the Okavango River basin, which is one of the principal sources of regional freshwater resource. The Okavango Delta is the terminal sink of the river water derives from two major tributaries, the Cubango in the west and the Quito in the east. It lies within a half-graben system, which has been shaped by rifting and faulting in the Kalahari Basin. The Okavango River, which originates in the tropical highlands of southern Angola, feeds the delta with approximately 11 billion cubic meters of water every year. Rainfall over the delta adds another 6 billion cubic meters to the annual water input. These large water inputs sustain 6,000 km2 of permanent swamps and another 6,000-12,000 km2 of seasonal swamps on the fan surface. Only 2% of the inflow leaves the delta as surface outflow, the rest is lost to evapotranspiration.
Figure 1: The Okavango Delta as seen from Terra Modis 28 July 2002. The Okavango Delta is situated between 21o30 E to 24o E longitude and 18o15 S to 20o45 S latitude approximately.
In all three riparian states, water demands are growing rapidly. The management of the Okavango Delta thus has to consider a very complex socio-economic framework as well as the uniqueness and the extremely high value of its ecology. The mathematical model of the delta will be used to predict the consequences of different management strategies and to analyse different development scenarios for the region. It may therefore contribute to the formulation of a sustainable development strategy for the delta.
The principal issues addressed within the framework of this project are the following:
The first issue is related to water quantity and distribution problem. As a basis for sound management of the resource, a robust, spatially distributed hydrological model is needed. So far, hydrological modeling of the Okavango Delta has only been operating with box-like discretization and the resulting models have been calibrated with surface outflow data only. In the framework of this project, a predictive model of the delta is to be developed, which couples surface and subsurface processes. Spatial resolution will be high enough to use water levels and remote sensing (air-borne and satellite) data to condition the model parameters. Aerial and satellite remote sensing-derived information will also be used for conditioning in a Bayesian parameter identification framework. Up-scaling techniques will be applied in order to take into account appropriately the sub-grid variability due to the complex small-scale patterns of channels, swamps and islands.
The second key question is related to water quality: Which hydraulic and biogeochemical processes ensure that the surface water in the delta is of low salinity, all the way down to the outflow, although every year some 300,000 tons of solutes are deposited in the system? It appears that the principal feature of the delta providing this ecosystem service is its extremely elongated land-water interface (some 10,000 km). A major part of this "coastline" is formed by islands, which act as solute sinks. Salt concentrations in the groundwater below the islands can be 1,000 times higher than in the surrounding swamp a few meters away. These high salinity contrasts suggests that locally, on the hot spots of salinization, density driven vertical flow may take place and transport the salt to the deep aquifers underlying the delta.
The third (and supportive to the other two) issue is related to hydrological modeling input parameters generation from airborne and space-based satellite imaging technology. Since the Okavango Delta is very large and hydrological modeling requires time-series spatial information, generation of reliable hydrological parameters need to be carried out by using aerial and satellite (optical and microwave) technology. Accurate Ground Control Points (GCPs) needed to be established for geo-referencing all input parameters using state-of-the-art Global Positioning System (GPS) technology (no GCPs are available for thedelta). A number of other issues such as spatial and temporal scale, accuracy, reliability, etc. need to be addressed before using derived parameters as inputs to the model.
Figure 2: Terrestrial photo of a portion of the Okavango Delta [Left: Permanent swamp; Right: Seasonal swamp]. Major topographic features of the delta are channels, swampy areas, islands and grasslands.
Note: Four major physiographic regions which comprise the Okavango Delta are: the panhandle (entry channel), the permanent swamp, the seasonal swamp, and occasional swamp (sand dominated occasionally flooded area). Photo by: T. Gumbricht.
Figure 3: Aerial image of a portion (4.32 km x 3.76 km) of the Okavango Delta. Major topographic features are islands (mid-grey to bright), salty land surface (brightest), channels (dark black), swampy areas (dark-grey to darkish)
GOALS:
The primary goal of this research project is to set up a spatially distributed, regional hydrological model for the Okavango Delta and to use it for the assessment of different water and land management scenarios. The project encompasses four principal areas of investigation:
The analyses will contribute to better understanding of the complex interaction of physical and biological processes shaping the Delta. It will demonstrate: (i) how hydrological parameters can be derived efficiently and automatically from aerial and satellite images, (ii) how modelling concepts can be applied to such a system, and (iii) how integration of spatial information from remote sensing and GPS into hydrological modelling can enhance model performance and reliability.
The resulting model is intended to serve as a basis for decision-making and regional water and land management. Thus, the project can contribute to the sustainable management of the regional water and land resources and to the preservation of the Okavango ecosystem.
RATIONALE:
The Okavango Delta is recognized as one of the world's last pristine wetland systems of outstanding ecological value. It is situated in a region of growing water scarcity and land requirements for development. Many potential users (farmers, cattle ranchers, domestic users, miners, industrial users) therefore compete for the scarce water and land resources. Thus, the sharing of the resources between these users and the ecosystem has to be managed. Efficient management includes a priori analyses of possible consequences caused by the planned actions. To this end, a model is required, which can predict the behavior of the system on the relevant spatial and time scales with reasonable accuracy.
NATURE OF PROJECT:
The project is carried out in cooperation with Botswana government's Department of Water Affairs (
DWA), the University of Botswana (UB); and two institutes of ETH Zürich: the Institute of Hydromechanics and Water Resources Management (IHW), and the Institute of Geodesy and Photogrammetry (IGP).
CONTACTS:
Peter Bauer, Prof. Wolfgang Kinzelbach, IHW, ETH Zürich; Tshito Babusi, DWA, Botswana; Krishna Talukdar, Emmanuel Baltsavias and Prof. Armin Grün, IGP, ETH Zürich.
Electronic contacts:
SUPPORTED BY:
ETH Zürich
In Collaboration With:
Botswana Government's Department of Water Affairs (DWA),
University of Botswana (UB), Okavango Research Group at the University of the Witwatersrand, Johannesburg, South Africa.