|
|
|
Water Resources
|
Reservoir Water Quality Monitoring Using Landsat TM Images and Indicator Kriging
Identification of excessive pollution regions indicator krging
In light of the environmental impact assessment, it often desires to know areas of excessive pollution and delineate their boundaries for further regulatory or remedy actions. Identification of excessive pollution areas is based on the concept of higher probability of exceedance, say 0.8 for example, with respect to a predetermined and meaningful threshold. In this study the threshold is chosen at CTSI = 55 and areas with exceeding probability greater tan 0.8 are considered excessive pollution regions.
Indicator is one of the many geostatistical estimation methods. Unlike the most commonly used ordinary kriging method which gives estimates of interested parameter at unsampled locations, indicator kriging estimates the global or local distribution of the interested parameter. Several cutoff values of the interested parameter, say Z is transformed into either 1 or 0, i.e.
| ij=( |
1 if zj£zc |
j=1,2,...,n |
|
| 0 if zj> zc |
Where Z c is the cutoff Value Z j is the observed values of Z, is the total number of samples and I j is called the indicator variable.
In geostatistics, the spatial variation of a random variable Z is characterized by a variogram function defined as:
gz(h) = 1/2[z(x+h)-z(x)]2
Where z(x) represents the value of Z at the location x. For indicatior kriging, variograms of indicator variables corresponding to different cutoff values must be established first and then global or local cumulative distribution function of Z can be estimated by using these indicator variograms. Readers are referred to Journal and Juijbgrets (1978) and Isaaks and Srivastave (1989) for details of ordinary and indicator kriging.
Using CTSI = 60 as cutoff value and applying indicator kriging to previously estimated CTSI (Figure 2), we identify the areas of excessive pollution for 8/31/1993 image (Figure 4). These are areas that may require remedy actions.
Figure 4 The spatial variation Indicator Kriging map on Aug. 31 1993.
Conclusions
- We have established empirical relations between the three water quality parameters and spectral parameters of TM images. Using these equations, we are able to investigate the spatial variation of reservoir trophic state. However, it is important to recognize that as more water quality samples being collected, modification of these empirical relations maybe necessary.
- Indicator kriging is a useful tool for identifying excessive pollution regions. It has great potential of applications in environmental regulation or remediation since this type of applications usually have clearly defined and meaningful threshold values that can be used as cutoff values in indicator kriging.
Acknowledgements
We gratefully acknowledge the support of Water Resources Department, Ministry of Economic Affairs (WRD - MOEA) in field data collection. This study would be possible without their support.
References
- Carlson, R.E. (1977). A trophic state index for lakes. Limnol. Oceanog., Vol. 22, No.2, pp. 361~369.
- She, F., X. Li, Q. Cai and Y. Chen (1996), Quantitative anlysis on Chlorophyll-a concentration in Taihu Lake using Thematic Mapper data, J. of Lake Sciences, Vol. 8, No. 3 o
- Isaaks, E. and R. Srivastave (1989). An Introduction to Applied Geostatistics. Oxford University Press, New York.
- Journal, A. and C.J. Huijbgrets (1978). Mining Geostatistics. Academic Press, New York.
- Lavery, P., C. Pattiaratchi, A. Wyllie and P. Hick (1993). Water quality monitoring in estuarine water using the Landsat Thematic Mapper, Remote Sensing of Environment, Vol. 46, pp. 268-280.
- Lillesand, T.M., W.L. Johnson, R.L. Deueil, O.M. Lixdstrom and D.E. Meisner (1983). Use of Landsat data to predict the trophic state of Minnesota lakes, Photogrammetric Engineering & Remote Sensing, Vol. 49, No 2, pp. 219~229.
- Tassan, S. (1993). An improved in-water algorithm for the determination of chlorophyll and suspended sediment concentration from Thematic Mapper data in coastal waters. Int. J. Remote Sensing, Vol.14 No. 6 pp. 1221~1229.
|
|
|
|
|
|
|
