Growing concerns about elevated nitrate levels in natural springs on the southern slope of the Irazu Volcano, Cartago province in Costa Rica, were the driving force for a multi-disciplinary study on the problem. The region is characterized by its high agricultural output and steep slopes. More than sixty springs located in the area are managed by a large number of local water authorities, ASADAS. The study focused on determining the main sources of nitrate pollution. Anthropological activities such as agricultural practices and domestic actions are found to contribute the most. A multivariate polynomial regression model was used with a large set of parameters. From the results it can be seen that human activities within the by law determined 200 meter radius protection zone around the springs, are most influential on high nitrate concentrations in the springs. Furthermore a stakeholder analysis, fieldwork, financial analysis of alternatives and farmer interviews were performed to produce a comprehensive list of recommendations for the various stakeholders. The recommendations are compiled to ensure a future with clean drinking water for all the citizens in the region.

This thesis investigates the efficacy of artificial intelligence (AI) models, particularly convolutional neural networks (CNNs) and U Net architectures, in reconstructing datasets with missing velocity data in river flow analysis. Optical flow and Particle Image Velocimetry (PIV) techniques have emerged as valuable tools for analyzing river flow patterns. Through a comprehensive literature review, CNNs, and UNet models are identified as promising tools for this task due to their ability to capture intricate patterns in datasets. The study compares the performance of the U-Net model against a statistics-based hydrological benchmark model, revealing the superior performance of the U-Net model. Furthermore, the analysis explores how the performance of AI models varies with differing quantities of missing data, by masking available data and comparing reconstructed values against the ground truth, highlighting the importance of data availability. Additionally, the study investigates the influence of spatial patterns in training data on model performance, including patchy versus random missing data in the field of view, simulating more datasets more likely available in reality. This clarifies the challenges encountered in predicting grid points under different training dataset conditions. Finally, the study identifies areas within the dataset that are particularly challenging to predict, shedding light on factors contributing to prediction errors. These findings underscore the potential of AI models in hydrological applications and provide valuable insights for future research in the field. Our findings show that U net is capable of reconstructing velocity fields from a river flow better than an average benchmark that uses the average values, with varying accuracy depending on input data. The average benchmark model had a relative error close to 0.2 in every instance, whereas the U-Net model showed relative errors ranging from 0.085 to 0.006. Errors from a patchy mask are ranging from 0.09 8 to 0.031.