I'm a GIS analyst and geospatial data scientist. I build spatial analysis workflows end to end: satellite remote sensing in Google Earth Engine, spatial SQL in PostGIS, modeling and statistics in Python and R, and interactive web maps. Most of my work is in environment, water resources, and planning.
29.8833° N, 97.9414° W · San Marcos, Texas
M.S. Geography · GPA 3.84
Background, current work, and what I'm looking for.
I started in geomatics engineering in Nepal, doing GNSS and UAV surveys and municipal GIS work, and moved into remote sensing and spatial data science during my M.S. in Geography at Texas State. These days most of my time goes to Google Earth Engine, PostGIS, Python, and R, applied to hydroclimatic analysis, suitability and flood-risk modeling, and land cover change.
I try to keep my work reproducible: public code, documented data sources, and a record of methodological decisions. The project pages below link to the repositories so you can check the work yourself. I'm looking for GIS Analyst, GIS Developer, and geospatial data science roles in environment, water resources, energy, and planning.
Grouped by area. The bolded items are the ones I use most.
Each project has a full write-up, and most link to code or a live app.
ERCOT's interconnection queue measured entirely in PostgreSQL/PostGIS: 1,714 projects geolocated from named interconnection points by trigram matching (47.6% match rate, published), county queue-pressure mapping, and cohort attrition showing only 28.8% of 2010–2020 entrants got built. Sequel to the solar siting project, and a data answer to its cost critique.
A Random Forest model of where utility-scale solar actually gets built in Texas. Distance to transmission matters about 9× more than sunshine, and a transfer test to North Carolina shows the grid-access threshold shifts with state policy. Spatial-CV ROC-AUC 0.92.
An interactive scrollytelling map of summer land surface temperature across all 65 Austin neighborhoods, measured in Google Earth Engine and joined to census income via a PostGIS spatial join. Canopy–heat r = −0.87.
The same rooftop-solar method on 1 m airborne LiDAR (Austin) and a 30 m open DSM (Kathmandu), quantifying how elevation-data resolution systematically biases solar estimates.
A live ArcGIS dashboard pairing real-time EPA AirNow air quality with a census-tract analysis of social vulnerability and monitoring gaps across Houston-Galveston-Brazoria. 399 of 1,595 tracts are both highly vulnerable and beyond neighborhood-scale monitoring.
A Markov chain–Cellular Automata model that learns 2010–2020 land-cover transitions and forecasts where San Antonio will grow by 2030 (Figure of Merit 0.85).
A custom ArcPy tool combining demographic need with transit service areas to find high-need neighborhoods underserved by bus stops in Portland, OR.
A statistical test of how much vegetation (NDVI) cools the land surface across five land-cover types, with a full reproducible R workflow (Kruskal–Wallis, ANCOVA).
Pre/post-fire change detection of the 2025 Pacific Palisades Fire using Wyvern hyperspectral imagery and red-edge vegetation indices.
From municipal GIS in Nepal to hydroclimatic research in Texas.
Hydroclimatic data analysis in Python and R; long-term climate and water-resources research; building reproducible data-processing workflows.
GNSS and UAV surveys for river, terrain, and surface-water analysis; multi-temporal remote sensing of Fewa Lake water-extent change for flood-risk management.
Multi-criteria land-use suitability analysis; wildfire-risk model for Banke National Park; cadastral/DGPS land pooling; municipal thematic mapping for planning and reporting.
M.S. Geography, Texas State · B.S. Geomatics Engineering, Tribhuvan University.
Open to new roles
I'm open to GIS Analyst, GIS Developer, and geospatial data science roles in environment, water resources, energy, and planning. Email is the fastest way to reach me.