Monitoring Desertification based on Geographic Information System and Multi- Spctral and Multi-Temporal Satellite Data Case Study; Damghan Playa

Results and Discussion
Using aereal photos, visual interpretation, field working, six training area were defined (table 1) and changed area were showed (fig. 2).

Using various change detection methods and post classification comparison, changes were detected and finally the changes map were produced (fig.3,4). Then changed areas were masked from TM and ETM+ bands.


Fig.2: Map of visual interpretation of changed area


Comparing this curves indicate that by eliminating of changes effects correlation between bands have increased (toward negative) but this trend is much considerable about ETM+ bands. In general in ETM+ sensor correlation between thermal and reflective bands is lower than TM bands. Because may be in this sensor is enhanced and thermal band has a higher spatial resolution. Therefore, information content and role of thermal ETM+ band to completing of reflective bands is more obviously.

Based on this research we can conclude that;
  1. The main land cover / land use types and soil salinity changes can be detected.
  2. The TM and ETM+ thermal bands contain complementary information to the TM and ETM+ reflective bands and combination of them may provide a strong tool for classification and separating of marginal playa lands.
  3. By comparison correlation of thermal and reflective TM and ETM+ bands we can conclude that generally these correlation in ETM+ band is lower than TM bands, indicating more information of thermal ETM+ bands cause to higher spatial resolution (60?60m).
  4. Ideally change detection methods should involve data acquired by the sensor with the same spectral and spatial resolution. Nature of change detection is complex especially with remote sensed data. But from the obtained result we conclude that in detecting the nature some spatial lands. Difference between spatial resolution of thermal bands might not be so important and more detected spectral information relevant to the physical and chemical composition can be more important.
But in the same condition because higher resolution ETM+ thermal band may provide more useful information for soil studies


Fig. 3: Map of changes area between 1977-1988



Fig. 4: Map of changes area between 1988-2000


5- Further research is necessary for comparison between the information content of TM and ETM+ data with the same data of recording.

4. Conclusion
In this study desertified land in Damghan playa was mapped from 1977 and 2000. The overlay analysis of the two land cover maps revealed that there is an imbalance in the spatial distribution of desertified areas. The marginal part of playa is mostly changed and desertified. By comparison its eastern part is better covered by vegetation. The changed land in desertified land is closely correlated with changes in salt and water, but loosely change with vegetation in total. In addition to the need for tow accurately classified scenes the detection of changes and hence the monitoring capability and Landsat is dependent on change in the spectral characteristics of various habitats through times. The reflectance characteristics of land and water surfaces are influenced by a number of factors. Additional factors that can influence the change are reflectance between three different. Landsat provides imagery of the earth surface on a regular basis for this reason, it is ideally suited to monitoring or detecting change over time. In spite of this there appear to have been few attempts to incorporated Landsat data into environmental monitoring programs. It is suggested that there are two main reasons for this. first many of the changes that occur are in too small a scale to be detected and mapped using Landsat data. In such cases aerial photography is more suitable and there are many examples in the literature where conventional photogrammetric or photointerpretation methods have been used to study changes in the natural environment.

From the above results, we may conclude that there are some problems in relation to the resolution for a multitemporal analysis based on Landsat MSS and TM/ETM+ images, but these problems can be mainly solved by training the broad classes in MSS images and regrouping the finer TM/ETM spectral classes to the classes corresponding to MSS classes. These post classification processes are not only necessary to obtain the same meaningful classes in both MSS and TM/ETM+ classified images, but are also useful to increase the accuracy of classification and consequently the accuracy of change detection. Therefore we concluded that the Landsat MSS and TM/ETM+ images and GIS are useful tools for change detection. From the result of multitemporal analysis we concluded that some drastic land cover changes took places in the area in period 1977-2000. The TM/ETM+ satellite images and GIS are offering a valuable contribution to fulfilling the information need in the natural resources management in the marginal part of the playa.

References
  • Alavi Panah, S.K., 1997. Study of soil salinity in the Ardakan area (Iran) based upon field observations, remote sensing and a GIS. P. 292. Gent: University of Gent (Ph. D. thesis).
  • Allen, R. C., Peter, R.J., Moore, J.Mc, and Gue, L.J., 1996. Classification based image enhancement for mineral mapping and mine site study, 11th ERIM Thematic Conference of Geological Remote Sensing, Las Vagas, Nevada, USA, USA 27-29 February, 1996.
  • Anthony Gar-On Yeh and Xia Li, 1996. Urban growth management in the Pearl river delta: an integrated remote sensing and GIS approach. ITC J. 1996-1: 77-86.
  • Ashby, W.C., Vogel, W.G., Kolar, C.A. Philo, 1984. Productivity of stony soils on strip mines. Proceeding of Erosion and productivity of soils containing rock fragments, Anaheima, Calif., 28 Nov.-3 Dec. 1982.
  • Goossenes, R. and E. Van Ranst, 1996. The use of remote sensing and GIS to detect gypsiferous soils in the Ismailia Province (Egypt), Proceeding of the International Symposium on Soils with Gypsum. Lleida, 15-21 September, 1996. Catalonia, Spain.
  • Jensen, J. R., 1983. Biophysical Remote Sensing . Annals of association of American Geographer, l (73): 111-132.
  • Lillesand, T.M., and Kiefer, R. W., 1994. Remote sensing and image interpretation, (4th edition), New York, John Willey and Sons.
  • Mather, P. M., 1987. Computer processing of remote sensing images, an introduction, John Wiley and Sons.
  • Todd, W. J., Gehring, D.G. and Hamman, J.F., 1980. Landsat wild mapping accuracy: Photogrametric Engineering and Remote Sensing 46 (4): 509-520.
  • Townshend, J. R. G. et al., 1989. Monitoring playa sediment transport system using thematic mapper data Advances in Space Research 9: 177-83.
  • Verma, K. S. and et al., 1994. Remote sensing technique for mapping salt affected soils. Intern. Journal of Remote Sensing, 15(9): 1901-1914.

Page 3 of 3
Previous