List of Tables
8. CONCLUSION AND RECOMMENDATIONS
8.3 RECOMMENDATIONS FOR FUTURE STUDIES
The research presented in this thesis, informed and influenced as it was by previous research described in the literature, has helped to provide a greater understanding of the surface waters and, by association, the wetlands of the Eastern Shores. At the same time, this research has raised new questions and presented new opportunities for further investigation.
8.3.1 Protocol validation
A number of assumptions were made during the development of the water mapping protocol. Some of these concerned the development of endmembers used to determine the mixing of water and vegetation, while others concerned the use of the Landsat Thematic Mapper. These assumptions suggest that a number of follow-up studies may be useful:
Verify the vegetation endmembers used in the protocol. The principal components analysis revealed that three endmembers would be sufficient to unmix the spectral data in the study area. These endmembers were shown to be water, woody and non-woody vegetation. In defining the non-woody vegetation endmember, it was assumed that the spectral reflectance characteristics of wetland and grassland vegetation are similar, and that a grassland endmember would adequately represent all non-woody vegetation. Measuring the actual reflectance characteristics of both grassland and wetland vegetation would verify whether this assumption was valid.
• Test whether the protocol works on other satellites. The water mapping protocol was developed using the spectral and spatial resolution of the Landsat Thematic Mapper. The protocol needs to be tested on data measured by satellites that have different spatial and spectral characteristics. This will provide an indication of the transferability of the protocol.
In addition to testing the influence of the satellite platform and endmember selection on the water mapping protocol, the inputs and outputs of the protocol can also be tested. In particular, it would be useful to examine the following:
• Use groundwater data to test the validity of the protocol. A link has already been established between cumulative rainfall and surface water extent. There is a need to extend this link to groundwater levels. The establishment of such a link would allow groundwater levels to be used in place of rainfall for verifying the protocol outputs.
Examine the effects of fire scars on the protocol. Analysis of fire scars in the study area revealed that these scars have spectral similarities to water. The effect of fire scars on mapping accuracy needs to be determined and ways of overcoming these effects need to be incorporated into the protocol.
Determine the applicability of the protocol to other areas of Maputaland and South Africa. The water mapping protocol was developed using the environmental conditions present on the Eastern Shores of Lake St Lucia. The accuracy of the water mapping capability of the protocol still needs to be tested for other areas of South Africa.
8.3.2 Relating water distribution and concentration to wetlands
The aim of this research was to map the surface water of the Eastern Shores. This has resulted in a series of water concentration maps depicting the spatial and temporal distribution of surface water in the study area. The availability of these maps allows a number of follow-up studies to be conducted:
Comparison between water concentration and vegetation. The water concentration maps indicate the distribution of water, usually in the presence of vegetation. A useful study would be to relate the different water concentration levels to wetland vegetation type, thereby allowing vegetation and wetland distribution maps for other regions to be produced based on the relationships established between water concentration and vegetation.
• Relate water concentration to wetland hydroperiod. Wetlands can be defined in terms of the hydroperiod, which describes the pattern of flooding in the wetland.
The series of maps depicting water distribution and concentration that were
produced in this study provide an indication of the hydroperiod for the whole of the study area. Linking these patterns of inundation to actual hydroperiod definitions will allow the different types of wetlands on the Eastern Shores to be classified on the basis of both physical and biological characteristics.
8.3.3 Modelling
The development of a water mapping protocol, which uses an established link between surface water and cumulative rainfall, allows the modelling of different scenarios to be undertaken. These include:
Investigate the effects of global warming-induced changes in rainfall patterns on the distribution of surface water. Most studies indicate that global warming will lead to changes in the world's rainfall. The effects of changing rainfall on the Eastern Shores, whether increasing or decreasing, can be examined by utilising the relationship that has been established between cumulative rainfall and surface water distribution.
Examine past patterns of surface water distribution based on historical rainfall patterns. By knowing the current relationship between rainfall and surface water distribution, past patterns of water distribution can be deduced from historical rainfall patterns. These rainfall patterns can either be obtained directly (i.e. from rain gauge records) or be inferred from proxy indicators like tree rings or sediments.
Examine the effects of removing the Eastern Shores plantations. The removal of the 5600 ha of commercial pine plantations, and the consequent reduction in evapotranspiration., is likely to have a measurable impact on the hydrology of the region. These impacts can be investigated by running the protocol without a forest mask, and then using topography and existing water distribution patterns to predict where surface water will reappear following the removal of plantations.
Examine the effects of sea level rise on surface water. Rising sea levels, induced by global warming, are likely to cause higher groundwater levels on the Eastern Shores. If it were possible to establish a link between groundwater levels and surface water distribution (as was done between rainfall and surface water in this study), then the protocol could be used to examine the effects of sea level rise on the surface water of the Eastern Shores.