Final Thoughts and Conclusions

Remotely sensed data, especially satellite imagery represents an extremely important and powerful tool for socio-economic and environmental research. It allows researchers to gather data on the biophysical context of social and environmental phenomena systematically and efficiently. Many social science and environmental science researchers have already begun using remotely sensed data to determine a number of land cover and climatic features. However, challenges remain in the fusion of remote sensing into social and environmental research, leaving remotely sensed databases vastly under-used. The case study presented here can be used to discuss the questions presented earlier in the paper. What level of sophistication in data analysis is necessary and realistic for social scientists? How can integration of remotely sensed data into mainstream research be improved?

While some social and environmental researchers may not have backgrounds or expertise in remote sensing technology, the use of data derived from these sources should not be out of reach. Data processing can be simple with the right tools, however, social scientists should be concerned with the methodology used to extract such information as vegetation change, particularly in forested environments. The results of this study in India indicate that simple methods, such as NDVI, can yield sufficient vegetation change information for many purposes in environmental research. However, areas of high vegetation abundance, such as forests, did yield less accurate vegetation data using NDVI than SMA. When this information was applied to research questions on land use change, the measure of percent live cover generated from sophisticated methodologies, such as SMA, seemed to be more easily integrated into meaningful variables than the index numbers of NDVI. Additionally, SMA had the added benefit of simultaneously generating soil fraction data as well as other land cover endmembers. This additional data was useful in evaluating phenomena such as desertification and forest change. And finally, when subtle change in forests are a focus of a study, the differences between NDVI and SMA become greater. Therefore, the sophistication of the methodology used to derive vegetation data should be decided based on the type environment studied and the types of changes that are important to observe.

The integration of derived data into research projects can be improved through a variety of uses as well as additional inputs. By itself remotely sensed data can be used to generate quantitative conclusions on the extent and trends in environmental changes, such as desertification, deforestation, and urbanization. However, these conclusions are strengthened by additional data sources which might help explain the observed phenomena and verify conclusions. These additional sources could be fieldwork, socio-economic statistics, census information, precipitation data, and even higher resolution remotely sensed data. While challenges exist, the possibilities are exciting and should warrant the effort necessary to incorporate this data into environmental and social science research.

Back to Index

For more information please contact Laurel_Firestone@yahoo.com