Ysagar-hub/KerichoClimateRiskpred-
Domaine:
agricultureclimateType de record:
software
Créateur:
Ysa
Hôte:
Kericho Multi-Hazard Climate Risk Model is a Python tool that uses NASA POWER weather data to assess agricultural climate risk in Kericho, Kenya. It detects key hazards, builds a Climate Risk Index, estimates crop loss and economic impact.
Author: Sagar Kumar
KerichoClimateRiskpred is a Python project that estimates climate risk for agriculture in Kericho, Kenya using daily weather data from the NASA POWER API (2010 to present). The script identifies common weather hazards, combines them into a Climate Risk Index (CRI), and uses that index to approximate crop loss and economic impact. It also labels each year into a simple risk category (low/medium/high), calculates an example insurance premium, and trains a Random Forest model to predict yearly risk levels. Finally, it produces graphs showing CRI trends, estimated losses, and hazard importance.
Location and data source
Location: Kericho, Kenya
Latitude: -0.367
Longitude: 35.283
Data source: NASA POWER (daily data for a single point)
Variables used:
T2M (average temperature)
T2M_MAX (maximum temperature)
T2M_MIN (minimum temperature)
PRECTOTCORR (precipitation)
WS2M (wind speed)
RH2M (relative humidity)
What the script does (step-by-step)
Downloads daily weather data from NASA POWER.
Cleans the dataset by replacing -999 with missing values and filling gaps.
Creates hazard indicators:
heavy rain (precipitation > 50 mm/day)
dry day (precipitation 32°C)
high wind (wind speed > 6 m/s)
Detects hail events using a rule-based proxy (rain + cold minimum temperature + wind + high humidity).
Builds a daily CRI score using weighted hazards, then aggregates it to an annual CRI (scaled 0–100).
Estimates crop loss percentage and converts it into economic loss using farm area and value per hectare.
Assigns a yearly risk class (LOW, MEDIUM, HIGH) and calculates an example insurance premium.
Trains a Random Forest model using annual hazard totals to predict the yearly risk class.
Saves results to CSV files and creates plots for CRI, loss trends, and feature importance.
How to run
Install dependencies:
pip install -r requirements.txt
Run the script:
python weatherrisk0.py
Replace weatherrisk0.py with whatever you saved the file as.
Output files
CSV outputs: …