Projects
1. ANN for downscaling Groundwater model - Repo, Poster,
This is my key PhD project where I have developed a scheme for downscaling key metrics for Groundwater (GW) Models uisng Remote Sensing (RS) and other modeling data via Artificial Neural Networks (ANN) approach. Conventionally, Groundwater Storage changes are required to be quantified using Groundwater Monitoring wells on the ground which are not able to measure such chnages on a consistent basis and are expensive to maintain. Satellite Remote Sensing based datasets such as from GRACE satellite provide harmonic coefficients which can be used to compute the Total Water Storage Changes through which Groundwater Storage Changes can be computed through rigorous analysis, althoigh such measurements still suffer from inaccuracies and coarse resolution of ~300 km which hinder accuarate analysis of GW storage chnages at a local level. We use GRACE satellite data along with Precipitation, Temperature, Evapotranspiration and SOil Moisture and other relevant data as input and Storage changes as Label data to improve the GWS estimates from satellite and to create better estimates for future as well. The work has been presented at key conferences Presentation for ANN approach in Grounwater Problems.
2. Glacier - Presentations, Repo,
During my undergraduate study, I led a research that analyzes and quantifies the mass mass balance for a key Himalayan glacier using a data driven approach. Glaciers are key groundwater resource and many Himalyan glaciers have undergone melting in recent decades. However ground based for monitoring health of such remote Himalayan glacier are few since these methods are laborious and expensive. Remote Senisng methods often used for studying glacier changes also suffer from measurements and processing error and lack of ground based validation. We used a holistic data driven approach to understand the different factors affecting the glacier health such as area and velocity chnages and provide a accurate estimate of the glcaier mass balance. The peer reviewed work has been extnsively followed and cited by the scientific community.
3. Computer Vision - Project files Classroom Projects
Used openCV library to work on Classroom projects based on Phottogrammetry, Computer Visison tasks such as Homographic Transformation, Image Mosaicking, etc
4.Urban Waste Monitoring and Classification with High Resolution Remote Sesning data - docs
This study focused in using GIS operations in an open sourec softwares and common classification operations such as