Komondi/Modelling-fire-frequency-in-Kenya

# Modelling-fire-frequency-in-Kenya This repository contains the data files and code used in the publication titled "A Bayesian Model for Predicting Fire Frequency in Kenya." The objective of the study was to create and evaluate a Bayesian model to predict fire frequency in Kenya. The repository includes code used to preprocess the data and combine it into the final form. It also includes simulation scripts used to generate the datasets used in the study. ## Repository Structure The repository is organized into the following folders: - **modis_data**: This folder contains CSV files of daily fire hotspot data downloaded from MODIS. The data spans the period from 2000 to 2020. - **fire**: This folder contains CSV files of fire frequency by month from 2000 to 2018. - **rainfall**: This folder contains GeoTIFF files of precipitation data in Kenya from 2000 to 2018. - **tmin**: This folder contains GeoTIFF files of minimum temperature data in Kenya from 2000 to 2018. - **tmax**: This folder contains GeoTIFF files of maximum temperature data in Kenya from 2000 to 2018. - **climate**: This folder contains merged CSV files of rainfall, fire, minimum temperature, and maximum temperature data matched by coordinates. The merged data is used for analysis. - **simulated_data**: This folder contains CSV files of data simulated during the study. These simulated datasets are used for model evaluation. - **model_results**: This folder contains model results on the simulated datasets. The results provide insights into the performance of the Bayesian model. ## Code Scripts The repository also contains several R scripts that perform various tasks: - **1import_merge**: This script contains code to import and merge climate and fire data. It processes the raw data files and combines them into a format suitable for further analysis. - **2data_aggregate**: This script merges data in the climate folder into a single file to be used in the analysis. It aggregates the data from …