Krishna
  • About
  • Skills
  • Experience
  • Projects
  • Education
  • Research

Krishna C. Mummadi

Geospatial Analysis • Remote Sensing • Earth Observation • GeoIntelligence

About Me

I am a graduate student in Geographic Information Systems (GIS) at the University of Maryland Baltimore County, with a research focus on Earth observation, remote sensing, and geospatial analysis. My work centers on using satellite imagery, spatial data, and reproducible workflows to study environmental and socio-environmental processes.

I am particularly interested in applying computational methods, Geospatial AI, and programming automation to analyze large geospatial datasets and support research in environmental monitoring and urban planning.

Technical Skills

Programming & Data Analysis

Python (Arcpy, Pandas) R MATLAB JavaScript HTML / CSS C / C++

GIS & Remote Sensing Tools

ArcGIS Pro & Online QGIS Google Earth Engine ENVI Pix4DMapper GEDI Hyperspectral Data ESRI Field Maps Leaflet AutoCAD

Database & Management

PostgreSQL / PostGIS Git Jupyter Notebooks NVivo Quarto raster / sf / terra tidyverse netCDF / HDF

Experience

GIS Fellow

June 2025 – Present

AGU's Thriving Earth Exchange Cohort • Baltimore Green Justice Workers Cooperative

  • Coordinate GIS analysis and mapping for 12 Baltimore-based climate organizations focused on climate and environmental programs.
  • Create StoryMaps, infographics, and data visualizations, and deliver GIS workshops to build community capacity.
  • Perform geospatial and remote sensing analysis to produce datasets and visualizations supporting grant proposals and stakeholder engagement.

Graduate Research Assistant – Geospatial AI & LLM

Sept 2025 – Present

University of Maryland Baltimore County • Baltimore, USA

  • Developing a workshop on Geospatial Analysis with Generative AI, integrating LLM-based tools with R, Quarto, and GIS workflows.
  • Organizing workshops on text and image analysis using Generative AI for qualitative social science research.

Graduate Research Assistant – Housing & Policing

Nov 2024 – Present

University of Maryland Baltimore County • Baltimore, USA

  • Conducting GIS-based research using ESRI ArcGIS and NVivo to generate spatial insights and thematic analysis on urban housing interfaces.
  • Contributing to a research paper on the intersection of urban housing and policing.

GIS Mapping & Data Intern

Feb 2025 – May 2025

American Dog Society • New York, NY

  • Extracted, cleaned, and analyzed geospatial data on pet facilities using Python to construct neighborhood-level metrics.
  • Developed interactive web maps via JavaScript, HTML, and ArcGIS Online.

GIS Senior Process Executive

Apr 2022 – Aug 2024

Infosys • Hyderabad, India

  • Developed Python-based automation tools with documented workflows, reducing manual GIS processing by 40%.
  • Led QA/QC analysis on Apple Maps’ POI data, improving navigation accuracy. (Awarded I-Star Incentive 2023)

GIS Intern

Jun 2021 – Aug 2021

Dronah • Jaipur, India

  • Automated GIS map generation for the Jaipur World Heritage City Tourism Plan using Python.

Planning Intern

Feb 2021 – May 2021

All India Institute of Local Self Governance • Hyderabad, India

  • Prepared Detailed Project Reports focusing on Sustainability and Waste Management for Gram Panchayats.

GIS Data Analyst

Aug 2020 – Feb 2021

Pure Earth Foundation • Hyderabad, India

  • Conducted on-site GPS surveys for forest land claims and mapped specific claimant polygons via ArcGIS/QGIS and LiDAR processing.
  • Led a team of 20 surveyors in GPS data collection, improving field accuracy by 25%.

Projects

Baltimore CVI satellite thermal imagery
Live Map StoryMap

Baltimore Climate Vulnerability Index

ArcGIS Online • Leaflet.js • JavaScript • FEMA NFHL • NLCD • HTML/CSS

  • Built a weighted composite index scoring every Baltimore census tract 0–100 across four indicators: Heat Severity (30%), Flood Exposure (25%), Canopy Deficit (25%), and Socio-Economic Sensitivity (20%).
  • Processed raster and vector data in ArcGIS Online using Zonal Statistics, Overlay, Summarize Within, and Min-Max Normalization across all inputs.
  • Developed for a GeoAI workshop — includes an interactive Leaflet map and an ArcGIS StoryMap presenting the full methodology and findings.
Transit accessibility satellite aerial
GitHub

Transit Accessibility of New Residential Development

R • Quarto • PostgreSQL/PostGIS • Montgomery County

  • Built reproducible ETL pipelines integrating parcel, boundary, and transit datasets into a PostGIS database.
  • Designed a normalized schema with spatial SQL queries to evaluate transit accessibility for 6,300+ new parcels.
  • Produced tract-level analyses to support equity-focused regional planning.
NDVI green space accessibility satellite imagery
GitHub

Green Space Accessibility and Demographics Analysis

R • OpenRouteService • ggplot2 • Leaflet • Baltimore

  • Conducted isochrone analysis using APIs for geospatial evaluation of park accessibility.
  • Analyzed walkability and demographics to identify park access disparities.
  • Created interactive Leaflet maps integrating ArcGIS to visualize underserved areas.
NASA PACE ocean color hypoxia satellite data
GitHub

Gulf Hypoxia Zones Assessment

Python • Jupyter Notebooks • PACE Satellite Data

  • Explored linkages between PACE ocean color and bottom-water hypoxia in the Gulf of Mexico (FISH-PACE Hackweek 2026).
  • Collated and analyzed in-situ dissolved oxygen (DO) datasets with PACE satellite imagery to understand biogeochemical relationships.
Baltimore vacant properties spatial analysis map
GitHub

Spatial Analysis on Baltimore Vacant Properties

R • ArcGIS • PostGIS • Baltimore Open Data

  • Integrated public and crowdsourced spatial datasets to analyze the distribution and clustering of vacant properties across Baltimore City.
  • Applied spatial statistics and PostGIS queries to identify neighborhood-level patterns and correlate vacancies with socio-economic indicators.

GeoAI & LLM Integration Experiments

Ongoing

R • Python • Generative AI • Quarto

  • Developing and testing workflows that integrate large language models into geospatial analysis pipelines using R and Python.
  • Exploring LLM-assisted qualitative coding, text analysis, and image interpretation for social science and environmental research.

Education & Credentials

Academics & Certifications

Masters in Geographical Information Systems (GIS)

May 2026

University of Maryland Baltimore County • GPA 3.82

Focus: Spatial Dataset Dev, Remote Sensing Apps, Satellite Image Analysis, Critical & Ethical Mapping.

Bachelor of Technology in Urban & Regional Planning

May 2020

Jawaharlal Nehru Architecture and Fine Arts University • India

Focus: Remote Sensing, AutoCAD, Spatial Data Analysis, Surveying, Photogrammetry.

Certifications

  • NASA ARSET: Hyperspectral Data for Land Cover Applications
  • C.E.D.: Advanced Certificate in Geoinformatics
  • QGIS Training: Central University of Karnataka
  • Brandefense Academy: OSINT — Visual Intelligence (VISINT) & Geospatial Intelligence (GEOINT) Apr 2026

Professional Footprint & Leadership

Conferences & Workshops

  • Esri FedGIS (Feb 2026)
  • FISH Pace Hackweek (Jan 2026)
  • TUGIS (Aug 2025)
  • Workshop Presenter — Geospatial Analysis & GeoAI Integration (Apr 2026) StoryMap Map

Leadership & Affiliations

  • Vice President - ASPRS (UMBC Student Chapter)
  • Student Member - Maryland Society of Surveyors & American Planning Association

MTIP Grant Recipient: Supporting GIS fellowship at BGJWC.
I-Star Incentive Award (2023): Outstanding contributions at Infosys.

Research Direction

I am preparing for doctoral study or research-focused roles in geospatial science, with an emphasis on remote sensing applications, environmental change analysis, and the integration of computational and AI-assisted methods in GIS research.

My current interests span Earth Observation, Land Use / Land Cover classification, urban climate vulnerability, environmental justice, and the use of Geospatial AI to support reproducible, scalable spatial analysis workflows.

© 2026 Krishna C. Mummadi. Built with precision for geospatial excellence.

Baltimore CVI — Full Methodology

7-Phase geospatial analysis workflow using ArcGIS Online, FEMA NFHL, NLCD & Census data

Phase 1

Data Collection

  • Administrative Boundaries: 2020 Census Tracts — Baltimore Metro Regional GIS Data Center
  • Thermal Analysis: Heat Severity USA 2025 (ArcGIS Image Service)
  • Flood Data: FEMA National Flood Hazard Layer 2024 (Flood Hazard Zones & Boundaries)
  • Vegetation Cover: NLCD Tree Canopy 2011–2021 (MRLC / ArcGIS Landscape Service)
  • Demographics: BaltimoreCity_Demographics & Poverty Status datasets
Phase 2

Heat Severity Processing (Raster → Vector)

  • Tool: Zonal Statistics to Table — Computes mean heat score per census tract from the raster.
  • Tool: Join — Joins the mean heat score table back to Census Tracts using Tract ID → TRACTICE10.
  • Output: Baltimore_tracts_heatseverity2 with average surface temperature per tract.
Phase 3

Flood Exposure Analysis

  • Tool: Overlay (Intersect) — Clips FEMA flood zones to census tract boundaries to create "flood pieces."
  • Tool: Filter — Retains only high-risk zones (FLD_ZONE: AE, A, VE, AO — ≥1% annual flood chance).
  • Tool: Summarize Within — Aggregates flood area (sq miles) per tract & calculates coverage percentage.
  • Formula: Flood_Percent = (Flood_Area_SqMi / Tract_SqMi) × 100
  • Output: Balt_tract_MHeat_flood — 0 = no flood risk; 100 = entire tract in floodplain.
Phase 4

Tree Canopy Deficit

  • Tool: Zonal Statistics as Table — Mean NLCD tree canopy % per census tract → Balt_tree_canopy.
  • Formula: Canopy_Deficit_Percent = 100 − MEAN_canopy (higher = less trees = worse).
  • Tool: Join Features — Joins canopy deficit to the heat + flood layer using Tract ID → CT10.
  • Output: Balt_Tract_Heat_Flood_Treecanopy
Phase 5

Socio-Economic Sensitivity

  • Tool: Join Features — Joins BaltimoreCity_Demographics to master layer using Full Census Tract ID → GEOID.
  • Social Vulnerability Formula (Arcade):
    ((100 − PNHT) + Percent_BelowPovertyLevel_All_p) / 2
  • Combines minority population index & poverty rate into a single 0–100 composite score.
  • Output: Balt_Tract_Heat_Flood_Treecanopy_final2
Phase 6

Min-Max Normalization (0–100 Scale)

VariableMinMaxNormalization Formula
Mean Heat1.094.02(x − 1.09) / (4.02 − 1.09) × 100
Flood Percent025.88(x − 0) / (25.88 − 0) × 100
Canopy Deficit32.84100(x − 32.84) / (100 − 32.84) × 100
Social Vulnerability49.6578.80(x − 49.65) / (78.80 − 49.65) × 100
Phase 7

Final Index Calculation

Weighted Composite Score (Arcade Expression)

(Heat_Norm × 0.30) + (Flood_Norm × 0.25) + (Canopy_Norm × 0.25) + (Social_Norm × 0.20)
  • Score 0: Safest tract — coolest, no flooding, maximum tree cover, high economic resource.
  • Score 100: Most vulnerable — hottest, flood-prone, bare canopy, low economic capacity.
  • Observed Range: 11.34 – 80.66 across all Baltimore census tracts.
Open Interactive Map View StoryMap