Mary Muthee

Geostatistical Modelling of Physical Intimate Partner Violence Among Women In Kenya Based on the 2022 Kenya Demographic Health Survey.

I analyzed 2022 Kenya Demographic Health Survey (KDHS) data to map the spatial variation of Intimate Partner Violence (IPV) at a 6km resolution using Binomial Classical methods and Monte Carlo Maximum Likelihood. Using binary logistic regression I identified the risk factors, selecting nine covariates based on literature. Higher IPV prevalence was found in Samburu, Bungoma, and Migori Counties. Key risk factors included low education, alcohol use by partners, poverty, lack of media exposure, urban residence, and being unmarried or employed.

Predicting the potential habitat areas for elephants in Africa using Maximum Entropy Algorithm under climate change

I modelled the current and future geographic distribution of African elephants under climate change using the MaxEnt model in R. Elephant occurrence data (GBIF) and bioclimatic variables (WorldClim) were used, with precipitation of the warmest quarter emerging as the strongest predictor. Results suggest a likely habitat shift toward southern Africa under climate change in future scenarios. This work was presented at the RIC2023 Conference and published in the abstract book.

Assessing Landform Changes in Kampong Plam Sugar Growing Areas

This project was aimed at assessing how land form changes in Kampong Speu Palm sugar growing areas between the year 2000 to 2024 using NDVI (Normalized Difference Vegetation Index), NDMI (Normalized Difference Moisture Index), MNDWI (Modified Normalized Difference Water Index),LST (Land Surface Temperature) and Land Use Land Cover(LULC) as the key indicators.

Monitoring Nitrogen Dioxide (no2) concentrations in Nyeri County from 2019 to 2022

The main objective was to monitor how Nitrogen Dioxide has been varying from the year 2019 to 2022 using Sentinel-5 Precursor (Sentinel-5P), which monitors a range of air pollutants. This is equivalent to the before lockdown, during lockdown and after lockdown period. From the study, there has been an increase in the No2 concentrations for the year 2021 and 2022 with Nyeri town and Mukurweini having the most concentrations while Mathira sub county has got the least concentrations for all the years.

Spatial Multi-Criteria Evaluation for Fire Disaster Response: A Case Study of Starehe, Nairobi

In this study, I used ArcGIS Network Analysis to support fire disaster response planning in Starehe Constituency, Nairobi. The analysis showed that the single fire station couldn't cover key areas like CBD, Kamukunji, Ngara, and Industrial Area within optimal response times. I performed shortest route, service area, and location-allocation analysis using data on roads, hospitals, population, land use, and fire incidents to identify ideal locations for new fire stations and improve emergency coverage.

Kenya Floods Interventions 2024 Dashboard

The dashboard visualizes flood response efforts in Kenya, tracking key metrics such as affected households, WASH distribution, food aid, and health outreach programs.

Use of GIS in disaster Risk Mapping: A Case Study of Earthquake in Izmir Turkey

I used GIS to map earthquake risk zones in Izmir, Turkey, following the 6.6 magnitude earthquake on 30 October 2020. Using spatial data, I identified high-risk areas, particularly around Izmir city and coastal zones near Seferihisar. The maps supported visualization of risk extent and informed disaster preparedness and response planning.

Interactive Web Map for Enhancing Tourism in Kenya: Attractions, Amenities, and Protected Areas

This web map shows Kenya's tourist attraction sites, protected areas(Game parks, National Reserves, National Parks and Forest Reserves), hotels and fueling stations. The main objective is to enhance tourism in Kenya. The map has a measure distance feature which the tourists can use to locate how far or near some of these amenities or sites are from their location.

Geo-Visualization of COVID-19 Vaccination Uptake in Kenya

This interactive map shows the rate of covid-19 vaccination uptake in Kenya per counties. It comprises of those who are partially vaccinated(one-dose) and fully vaccinated (two doses). Kenya Health Facilities have also been mapped .

Vegetation Classification Using LiDAR Point Clouds

I processed LiDAR point cloud data in R to generate Digital Surface Models (DSM) and Digital Terrain Models (DTM), from which Canopy Height Models (CHM) were derived. Using the lidR, terra, and raster packages, the CHMs were classified into three vegetation strata: trees, shrubs, and herbaceous layers.