An analysis of the sensitivity of soil erosion models within a geographical information system
Soil erosion is an important field of study due to the effects it has on the quality of soil and the environment. In the field of soil erosion modelling, the availability of data is an important factor that may influence the choice of model to be used. This is especially so in developing countries w...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
1998
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/7935/1/NorkhairIbrahimPFGHT1998.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Soil erosion is an important field of study due to the effects it has on the quality of soil and the environment. In the field of soil erosion modelling, the availability of data is an important factor that may influence the choice of model to be used. This is especially so in developing countries where data are scarce. Where data are available, their quality is seldom assessed, hence the level of reliability of modelled results is not known. The importance of availability and quality of data in any work involving GIS is well understood. Data for this study are obtained from Langkawi Island, Malaysia. Malaysia is a country where development is going on at a very fast rate. Activities such as clearing of forests for development have increased the risk of erosion, hence attention is now being given to this area so that threats and problems caused by it can be managed. This study analyses the sensitivity of two empirical models, the Morgan, Morgan and Finney (MMF) and Soil Loss Estimator for Southern Africa (SLEMSA) soil erosion models, to variations in their input variables by the use of geographical information systems. This is achieved by implementing the models in ERDAS Imagine Spatial Modeler wbile IDRISI is used to carry out most of the data manipulation, analysis and presentation. The study has shown that using the Spatial Modeler is a very efficient way of running the models especially when many model runs are involved. The results of these analyses show that one of the most important input variables is slope (the other being vegetation, soil type and rainfall). An analysis is carried out to find the best slope data set that can be derived from available topographic data. This is done by generating DEMs and slope images using three GIS software packages, namely the ERDAS Imagine-ARCIINFO combination, GRASS, and IDRISI. Pixel sizes of 5In, 20m and 50m are used in the analysis. The accuracy of the slope images is compared with values derived from a topngraphic map. It is found that slope data derived from ERDAS Imagine at 50m resolution are the most accurate and this data set is used in the subsequent analysis. Sensitivity analysis of the two models is carried out by using the one-at-a-time approach in which the model is run with the values of a single input variable being changed while keeping the other inputs at nominal values. Relative sensitivities of the model to variations in the input variables are deterntiued by assessing the confidence limits, bench marks and sensitivity parameters of the input variables. Generally, it is found out that the MMF model can be considered to be sensItive to three categories of input variables while the SLEMSA model is sensitive to two different categories. The results have shown that there are differences in the way the two models "react" to their various input variables, leading to the conclusion that due attention should be given to these inputs in terms of their levels of accuracy so that the results of soil erosion modelling can be interpreted in a more meaningful fashion. |
|---|