Course 31 - Global Climate Change and Vector-Borne Disease Risk Modeling

📖 Key Terms for Beginners

Before starting, make sure you understand these terms. They appear throughout all code blocks in this course.

Term	Full Name	What it means in simple words
Vector-Borne Disease	Vector-Borne Disease (VBD)	Infectious diseases transmitted by arthropods such as mosquitoes, ticks, and sandflies.
Environmental Suitability	Environmental Habitat Suitability	The degree to which environmental conditions satisfy the ecological requirements for vector survival and reproduction.
Thermal Performance Curve	Thermal Performance Curve (TPC)	The non-linear relationship between ambient temperature and vector biological traits like biting rate or pathogen incubation.
Extrinsic Incubation Period	Extrinsic Incubation Period (EIP)	The time required for a pathogen to develop inside a vector before becoming transmissible to a human host.
Environmental Exposure	Environmental Health Exposure Proxy	Satellite-derived measures of environmental conditions that favor vector habitat, distinct from direct disease occurrence.
Random Forest	Random Forest Classifier	An ensemble machine learning technique combining multiple decision trees to predict suitability from environmental covariates.
Spatial Autocorrelation	Spatial Autocorrelation	The tendency for geographical observations close to each other to have similar environmental or epidemiological values.
ERA5-Land	ECMWF ERA5-Land Reanalysis	A high-resolution global atmospheric dataset providing continuous climate estimates over land.
CHIRPS	Climate Hazards Group InfraRed Precipitation with Station data	A quasi-global rainfall dataset designed for monitoring precipitation extremes and climate trends.
LST	Land Surface Temperature	The radiative skin temperature of the Earth surface measured by infrared satellite sensors.
Ecological Niche Model	Ecological Niche Modeling (ENM)	A computational method mapping environmental conditions that correspond to species presence or suitability.
Confounding Variable	Confounding Variable	An unmeasured factor such as vector control measures or healthcare infrastructure that alters observed health relationships.

Introduction to GeoAI in Climate-Sensitive Health Modeling
Understanding Vector Biology and Thermal Performance Bounds
Multi-Sensor Climate Data Ingestion in Earth Engine
Preprocessing Temperature and Precipitation Covariates
Constructing Vector Physiological Suitability Proxies
Generating Pseudo-Presence and Background Samples
Training a Random Forest Environmental Suitability Model
Population Exposure Analysis with WorldPop Integration
Model Evaluation and Spatial Uncertainty Assessment
Ethical Communication and Limitations in Health GeoAI

Introduction

10 min

Global climate change is altering the geographic range and seasonal transmission dynamics of vector-borne diseases such as dengue, malaria, and chikungunya. This course introduces an advanced GeoAI framework that integrates ERA5-Land climate reanalysis, MODIS land surface temperature, and CHIRPS precipitation data to evaluate vector habitat suitability across tropical and subtropical regions. Students will combine remote sensing climate proxies with Random Forest classification in Google Earth Engine to predict environmental suitability and assess human population exposure.

GEE Asset ID

ECMWF/ERA5_LAND/MONTHLY_AGGR

Dataset Name

ERA5-Land Monthly Aggregated Climate Reanalysis

Study Area

Global Tropical and Subtropical Regions

AI Tool

DeepSeek (free, no API key needed)

🌍

What you will build: By the end of this course you will have a working GEE script that analyzes Global Climate Change and Vector-Borne Disease Risk Modeling, computes statistics, visualizes results on an interactive map, and exports data to Google Drive.

Setup: GEE + DeepSeek

5 min

You need two browser tabs open side by side throughout this course.

Open chat.deepseek.com -- free account, no API key needed

Open code.earthengine.google.com in a second tab

Create a new GEE script named Course31_Global Climate Change and Vector-Borne Disease Risk Modeling

💡

Workflow: Write a prompt in DeepSeek describing your task with exact dataset IDs and requirements. Copy the generated GEE JavaScript. Paste into GEE code editor. Click Run. Inspect results. Iterate.

Crafting the DeepSeek Prompt

15 min

The prompt is the core skill of this course. A good prompt includes the exact dataset ID, band names, study area coordinates, dates, preprocessing steps, and desired outputs.

📋 Main Prompt -- copy and paste this into DeepSeek

Act as an expert geospatial data scientist and spatial epidemiologist. Write a Google Earth Engine JavaScript script for modeling environmental suitability for disease vectors across global tropical and subtropical regions. Use ERA5-Land monthly reanalysis ('ECMWF/ERA5_LAND/MONTHLY_AGGR') for air temperature ('temperature_2m'), MODIS LST ('MODIS/061/MOD11A2') for land surface temperature ('LST_Day_1km'), and CHIRPS ('UCSB-CHG/CHIRPS/PENTAD') for precipitation ('precipitation'). Filter the data for the year 2022. Define a bounding box for tropical/subtropical regions between latitudes -30 and 30. Calculate annual mean temperature and total precipitation. Create rule-based physiological suitability thresholds for vector survival (e.g., temperature between 20C and 32C, annual precipitation > 500mm). Sample training points for pseudo-presence and absence, train a Random Forest classifier using ee.Classifier.smileRandomForest(100), and produce a continuously scaled suitability map. Include clear visualization layers and code comments explaining the epidemiological context and limitations.

✨

After pasting the prompt: Wait 20-30 seconds for DeepSeek to generate the full GEE script. Copy the entire code block it produces. If the output is cut short, type "Continue the code" to get the rest.

Ingesting and Preprocessing Multi-Sensor Climate Covariates

10 min

Paste the following code into your GEE code editor and click Run. Read the beginner notes below to understand what each line does.

Ingesting and Preprocessing Multi-Sensor Climate Covariates

var roi = ee.Geometry.Polygon([
  [[-120, 30], [-120, -30], [150, -30], [150, 30]]
], null, false);

var startDate = '2022-01-01';
var endDate = '2022-12-31';

var era5 = ee.ImageCollection('ECMWF/ERA5_LAND/MONTHLY_AGGR')
  .filterDate(startDate, endDate)
  .select('temperature_2m');
var meanAirTempC = era5.mean().subtract(273.15).rename('air_temp_c');

var modisLST = ee.ImageCollection('MODIS/061/MOD11A2')
  .filterDate(startDate, endDate)
  .select(['LST_Day_1km', 'LST_Night_1km']);
var meanLST = modisLST.mean().multiply(0.02).subtract(273.15).rename(['lst_day_c', 'lst_night_c']);

var chirps = ee.ImageCollection('UCSB-CHG/CHIRPS/PENTAD')
  .filterDate(startDate, endDate)
  .select('precipitation');
var totalPrecip = chirps.sum().rename('annual_precip_mm');

var predictorStack = meanAirTempC
  .addBands(meanLST.select('lst_day_c'))
  .addBands(totalPrecip)
  .clip(roi);

Map.centerObject(roi, 3);
Map.addLayer(predictorStack.select('air_temp_c'), {min: 10, max: 35, palette: ['blue', 'yellow', 'red']}, 'Mean Air Temp (C)');
Map.addLayer(predictorStack.select('annual_precip_mm'), {min: 100, max: 3000, palette: ['white', 'cyan', 'darkblue']}, 'Annual Precip (mm)');

🔑

Beginner note: This script defines a study region covering tropical and subtropical belts (-30 to +30 latitude). It extracts monthly mean air temperature from ERA5-Land, converts Kelvin to Celsius, calculates 8-day MODIS day land surface temperature, and sums pentad CHIRPS precipitation for 2022. The layers are combined into a single multi-band image stack for downstream environmental suitability modeling.

Machine Learning Suitability Modeling and Risk Stratification

10 min

Add this block after the previous code. It continues the analysis with additional processing and visualization.

Machine Learning Suitability Modeling and Risk Stratification

var tempSuitability = predictorStack.select('air_temp_c').gte(20).and(predictorStack.select('air_temp_c').lte(32));
var precipSuitability = predictorStack.select('annual_precip_mm').gte(500);
var pseudoSuitability = tempSuitability.and(precipSuitability).rename('suitability');

var presencePoints = pseudoSuitability.selfMask().stratifiedSample({
  numPoints: 400,
  classBand: 'suitability',
  region: roi,
  scale: 20000,
  geometries: true
}).map(function(f) { return f.set('label', 1); });

var absencePoints = pseudoSuitability.not().selfMask().stratifiedSample({
  numPoints: 400,
  classBand: 'suitability',
  region: roi,
  scale: 20000,
  geometries: true
}).map(function(f) { return f.set('label', 0); });

var trainingData = presencePoints.merge(absencePoints);

var trainingSamples = predictorStack.sampleRegions({
  collection: trainingData,
  properties: ['label'],
  scale: 20000,
  tileScale: 4
});

var classifier = ee.Classifier.smileRandomForest(100)
  .train({
    features: trainingSamples,
    classProperty: 'label',
    inputProperties: predictorStack.bandNames()
  });

var suitabilityMap = predictorStack.classify(classifier);

Map.addLayer(suitabilityMap, {min: 0, max: 1, palette: ['lightgray', 'orange', 'red']}, 'RF Environmental Suitability Proxy');

🔑

Beginner note: This script defines biological heuristic rules (temperatures between 20C and 32C and rainfall exceeding 500mm) to generate pseudo-presence and absence training locations. A Random Forest classifier with 100 decision trees is trained on the multi-sensor climate stack. The model classifies global environmental suitability into a continuous spatial indicator.

Export Results to Google Drive

5 min

Add this block at the end of your script to export the final map as a GeoTIFF to your Google Drive. After clicking Run, go to the Tasks tab in GEE and click Run next to each export task.

Export to Google Drive

Export.image.toDrive({
  image: suitabilityMap,
  description: 'Global_Tropical_Vector_Suitability_2022',
  folder: 'GEE_Vector_Health_Exports',
  fileNamePrefix: 'vector_suitability_rf_2022',
  region: roi,
  scale: 10000,
  crs: 'EPSG:4326',
  maxPixels: 1e9
});

🔑

Beginner note -- Export.image.toDrive(): This function does not export immediately. It creates a task in the GEE Tasks tab that you must manually start by clicking Run. The file will appear in your Google Drive folder (GEE_Courses) as a GeoTIFF that you can open in QGIS or ArcGIS.

Exercises Using DeepSeek Prompts

Homework

Use DeepSeek to extend the script for each exercise below. Write a follow-up prompt describing what you want to add, then copy the generated code and paste it into your GEE script.

Sensitivity Analysis of Thermal Bounds

Modify the temperature lower threshold from 20C to 16C in the suitability rule script.
Re-run the Random Forest classification to generate an updated suitability map.
Compare the change in total high-suitability surface area across subtropical margins.

Seasonal Dynamic Analysis

Select quarterly time intervals (Q1: Jan-Mar, Q3: Jul-Sep) for ERA5-Land and CHIRPS.
Compute quarterly mean temperature and precipitation stacks.
Map seasonal shift in suitability boundaries across the Sahel and South Asia.

Human Exposure Overlay with WorldPop

Load the WorldPop global unconstrained population count image collection for 2020.
Mask the population layer using high suitability zones (suitability equal to 1).
Calculate the estimated population living within highly suitable vector environments.

Spatial Autocorrelation and Model Validation

Perform a spatial split on training and test samples using geographical grid blocks.
Evaluate model accuracy metrics on geographically hold-out validation points.
Document how spatial autocorrelation might overestimate model performance.

Prompt Engineering Tips

Reference

To get the best results from DeepSeek for GEE tasks, always include these elements in your prompts:

Element	Example
Exact dataset ID	`ECMWF/ERA5_LAND/MONTHLY_AGGR`
Band names	"select band 'B4' for red, 'B8' for NIR"
Study area	"rectangle [-6.5, 33.0, -1.5, 36.0] North Morocco"
Date range	"filter from 2020-01-01 to 2023-12-31"
Output format	"print area in hectares, add a legend, export to Drive"
Error fixing	"The code gives error X. Please fix it."
Iteration	"Add a time series chart for each year 2018-2023"

🧠

Best practice: Start with the main prompt, run the code, then send follow-up prompts to add features one at a time. This gives you cleaner, more reliable code than asking for everything at once.

Global Climate Change and Vector-Borne Disease Risk Modeling

📖 Key Terms for Beginners

Contents

Sensitivity Analysis of Thermal Bounds

Seasonal Dynamic Analysis

Human Exposure Overlay with WorldPop

Spatial Autocorrelation and Model Validation