Integrating multivariate statistical analysis and geospatial method to study the sediment contamination in the Chao Phraya River network | |
| Author | Bhandari, Roshan |
| Call Number | AIT Thesis no.EV-22-11 |
| Subject(s) | Contaminated sediments--Thailand--Chao Phraya River Sedimentation and deposition--Environmental aspects--Thailand--Chao Phraya River |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Environmental Engineering and Management |
| Publisher | Asian Institute of Technology |
| Abstract | This research investigated the heavy metal and nutrient contaminations in the great Chao Phraya River network of Thailand using the multivariate statistical analysis and geospatial methods and explored their possible integration. Multivariate statistical analysis includes Principal Component Analysis and Canonical Correlation Analysis which were employed to find potential sources of sediment contaminants and the relationships with predictor variables respectively. The comparison of traditional regression model (Ordinary Least Square Regression) was conducted with Geographically Weighted Regression (GWR) model and shed light on the advantages on using GWR model. The suitability of conducing GWR model was analyzed explaining the spatial variation and the local relationships of contaminants with predictors and a suitable approach was identified in sediment study. Large pool of explanatory variables was used, which is novel in such type of study. PCA-GWR integration was executed to explore their benefits of combination. Pearson’s correlation reflected the decreasing degree of association between the contaminants on moving to the downstream in line with anthropogenic interferences. In general, PCA results revealed Cr and Ni potentially sourced from natural or lithogenic origin, Pb, Cd and Zn from urbanization and industrialization, Hg and As from the mining and industries, nutrients (TC, TN and TP) from domestic activities and agriculture. However, PCA of upper Chao Phraya showed considerable differences in the source of some of the contaminants to that of lower Chao Phraya supported by Pearson’s correlation, which could be attributed to the different proportion of land use types and the level of anthropogenic interferences. PCA-GIS based IDW interpolation of factor scores further demonstrated the lower Chao Phraya River with higher degree of contamination and association than upper Chao Phraya River. GWR model is advantageous over OLS model in explaining the spatial variation of sediment contaminants demonstrated by its higher R2 (over 2-fold time OLS R2 ), lower Akaike Information Criterion, and reduced spatial autocorrelation problem. Multivariate GWR model was with higher local explanatory power and performance than univariate GWR model. The performance of GWR model increased with the increasing level of urbanization i.e., the variability of contamination was explained well in the developed watershed that shows GWR is suitable in developed watersheds. GWR local parameters reflected that anthropogenic factor such as urban land, Population Density, Gross Provincial Product and industrial land were the most influencing predictors with higher local correlation for most of the heavy metals and nutrients. Apart from this, slope and rainfall as non-anthropic factors had strong influence on sediment contamination. An integration approach of GWR- linear regression model showed the relationship between the contaminants and land use predictors including PD in general becomes stronger with the increase in urbanization level except for agricultural land. This demonstrates that the presence of highly urbanized watershed weakens the impact of agriculture, while the low urbanized watersheds increase the influence of agriculture in sediment contamination. PCA-GWR integration approach generated quite consistent results for majority of the PCs like in comprehensive OLS-GWR approach, except for PC4. This integration approach could provide new perspectives to understand the data, and possibly generates valuable information for management and decision making. |
| Year | 2022 |
| Type | Thesis |
| School | School of Environment, Resources, and Development (SERD) |
| Department | Department of Energy and Climate Change (Former title: Department of Energy, Environment, and Climate Change (DEECC)) |
| Academic Program/FoS | Environmental Engineering and Management (EV) |
| Chairperson(s) | Xue, Wenchao |
| Examination Committee(s) | Ekbordin Winijkul;Virdis, Salvatore G.P. |
| Scholarship Donor(s) | Kurita Water and Environment Foundation, Japan |
| Degree | Thesis (M. Sc.) - Asian Institute of Technology, 2022 |