Loading...

Integration of Artificial Intelligence, Remote Sensing, and Field Data for Simulation of Chlorophyll-a Concentration in Gorgan Bay

FatemiHarandi, Mohammad Reza | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57478 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Khorashadizadeh, Farkhondeh
  7. Abstract:
  8. Chlorophyll-a, as an important component in evaluating water resources, has a significant impact on the status of water resources. This component is dependent on various quality parameters, such as phosphorus concentration, nitrogen concentration, turbidity, suspended solid concentration, temperature, pH, and dissolved oxygen concentration in water. To evaluate the quality status of surface water, it is necessary to estimate the exact concentration of chlorophyll-a in different temporal and spatial ranges. The objective of this study is to simulate and predict the concentration of chlorophyll-a in different temporal and spatial ranges using a combined application of machine learning models, remote sensing data, and field data. The study area is the Gorgan Bay and its adjacent rivers. Hydrological and quality data of the rivers, regional meteorological data, and satellite data of surface water temperature and dissolved organic carbon (POC) are considered as input and chlorophyll-a concentration as output for the machine learning model. The performance of the machine learning model is evaluated using various statistical indices. To better understand the effective parameters in chlorophyll-a concentration, a sensitivity analysis is performed using the three indices of variable importance (PVI), P-Value, and Gini analysis. A group sensitivity analysis is also performed on the input parameters, comparing the results obtained from executing the random forest model for each station's field data and satellite data separately. The results show that the random forest model performs better than the Least Square Support Vector Regression (LSSVR) model. Additionally, the data of surface water temperature and dissolved organic carbon in the Gorgan Bay have a greater impact on chlorophyll-a concentration in the Gorgan Bay area compared to meteorological data and data from the adjacent rivers. The probable reason for these results is the relatively large distance between the data collection stations on the rivers and the Gorgan Bay. Using the developed model in this study, different scenarios of water quality improvement and climate change can be evaluated, and it will contribute to better management of water quality in the Gorgan Bay
  9. Keywords:
  10. Chlorophyll-A ; Remote Sensing ; Remote Sensing Data ; Machine Learning ; Gorgan Bay ; Field Data

 Digital Object List

 Bookmark

No TOC