Assessing water body ecological indicators using time series analysis and physics-based modeling approaches.
Bhattacharjee, Natalia Valeryevna
MetadataShow full item record
This dissertation goal is to present a comprehensive time series analysis along with a physics-based model of a stream that can yield expanded ecological data over time which can be further studied with time series analysis. The time series component was primarily developed using available data from a windrow composting operation and runoff collection pond. Descriptive time series analysis (spectral analysis) and detailed recurrent neural network modeling (which requires training and validation data) were performed. Physical stream models and subsequent environmental flow analysis models were developed as aids for water resource management in the Ocmulgee and the Middle Oconee Rivers where we develop a platform useful for evaluating trade-offs between ecological impacts and economic development. We then selected one of the four environmental flow conditions for time series analysis. The methodology and approach set in this work may be adapted to inform environmental flow analyses in other study sites.