Skip to content

Automatic Weather Station Suitability Map

Automatic Weather Station Suitability Map – Province of Bukidnon

Overview

A GIS-based site suitability analysis identifying the most suitable locations for establishing Automatic Weather Stations (AWS) across the Province of Bukidnon, Philippines. Using the Analytic Hierarchy Process (AHP) as the multi-criteria decision framework, five spatial criteria were weighted and combined to produce a final suitability map classified into four levels — from Not Suitable to Highly Suitable — covering all municipalities of Bukidnon. The output is intended to support data-driven decision-making for weather monitoring infrastructure placement.

Study Area: Province of Bukidnon, Philippines — all municipalities
Role: Solo project
Status: Completed
CRS: PRS 92 / Zone 5 (EPSG:3125)

Disclaimer: Subject for site validation.

Methods & Tools

Data Sources

Dataset Source
Land Use / Land Cover (10m, 2024) Esri | Sentinel-2 Land Cover Explorer
Digital Elevation Model (DEM) USGS SRTMGL1 003 via Google Earth Engine
Road Network OpenStreetMap Philippines via Geofabrik Download Server
River & Water Bodies OpenStreetMap Philippines via Geofabrik Download Server
Existing AWS Locations PAG-ASA Mindanao (Philippine Atmospheric, Geophysical and Astronomical Services Administration)
Administrative Boundaries GADM (Global Administrative Areas) via QGIS Plugin

Suitability Criteria & Scoring

Criteria scoring was based on Felipe et al. (2024) as adopted by Dawis et al. (2025):

Criteria Highly Suitable (3) Moderately Suitable (2) Marginally Suitable (1) Not Suitable (0)
Slope ≤ 5% 5–15% 15–25% ≥ 25%
Land Use / Land Cover Class 40 Classes 20, 30, 60 Class > 100 Classes 0, 50, 80, 90
Distance to Water Bodies ≥ 1 km 0.5–1 km 0.25–0.5 km ≤ 0.025 km
Distance to Roads 0–0.7 km 0.7–1.2 km 1.2–2.2 km ≤ 0.3 / ≥ 2 km
Distance to Existing AWS ≥ 9 km 5–9 km 3–5 km ≤ 3 km

AHP Weights

Criteria weights were derived using the Analytic Hierarchy Process:

Criteria Weight
Slope 0.220
Land Use / Land Cover 0.223
Distance to Roads 0.185
Distance to Water Bodies 0.166
Distance to Existing AWS 0.205

Final Suitability Index Formula (Raster Calculator)

( [Suitability] Slope    × 0.220 ) +
( [Suitability] LULC     × 0.223 ) +
( [Suitability] Road     × 0.185 ) +
( [Suitability] Water    × 0.166 ) +
( [Suitability] AWS      × 0.205 )

Processing Workflow

The full analysis was conducted in QGIS and Google Earth Engine following these stages:

Stage 1 — Data Acquisition

  • Downloaded Sentinel-2 LULC GeoTIFFs (2024) from Esri Land Cover Explorer and uploaded to GEE as image assets
  • Extracted DEM (SRTMGL1) for Bukidnon from GEE and exported to Google Drive
  • Downloaded OSM road, waterway, and water body shapefiles from Geofabrik
  • Uploaded Bukidnon municipality boundary shapefile as a GEE and QGIS asset

Stage 2 — Preprocessing in QGIS

  • Clipped road, waterway, water body, and existing AWS layers to the Bukidnon municipality boundary
  • Converted water body polygons to lines for proximity analysis
  • Merged waterway and converted water body layers into a single water network layer
  • Converted DEM elevation to slope (degrees) using QGIS Raster Analysis > Slope
  • Reprojected all raster layers (Slope, LULC) and vector layers (Water, Road, AWS) to PRS 92 / Zone 5 (EPSG:3125)

Stage 3 — Rasterization & Proximity

  • Rasterized road and water network vector layers to 30m resolution
  • Rasterized existing AWS point layer to the same extent and resolution
  • Generated proximity (raster distance) maps for road, water, and AWS rasters using QGIS Raster Analysis > Proximity

Stage 4 — Reclassification

Each criteria layer was reclassified to a 0–3 suitability score using QGIS Raster Analysis > Reclassify by Table:

  • Slope — 0–2.9° → 3, 2.9–8.5° → 2, 8.5–14° → 1, 14–90° → 0
  • LULC — Crops (5) → 3, Bare Ground (8) / Rangeland (11) → 2, Trees (2) → 1, Water / Flooded / Built / Clouds → 0
  • Distance to Water — 0–25m → 0, 25–1000m → 1, 1000m+ → 3
  • Distance to Roads — 0–300m → 3, 300–700m → 3, 700–1200m → 2, 1200–2200m → 1, 2200m+ → 0
  • Distance to AWS — 0–3000m → 0, 3000–5000m → 1, 5000–9000m → 2, 9000m+ → 3

Stage 5 — AHP Weighted Overlay & Final Classification

  • Applied the AHP formula in QGIS Raster Calculator to produce a continuous Suitability Index
  • Reclassified the index into four final classes: Not Suitable (0–0.75), Marginally Suitable (0.75–1.50), Moderately Suitable (1.50–2.25), Highly Suitable (2.25–3.00)
  • Styled and exported the final map with municipal boundaries, legend, north arrow, and scale bar in QGIS Print Layout

Tools Used

Tool Purpose
QGIS All geoprocessing — clip, reproject, slope, rasterize, proximity, reclassify, raster calculator, map layout
Google Earth Engine DEM and LULC extraction, asset management, and export to Google Drive
Claude AI Code assistance, documentation, and project write-up

Suitability Legend

Color Class
🟥 Red Not Suitable
🟧 Orange Marginally Suitable
🟨 Yellow Moderately Suitable
🟦 Blue Highly Suitable

References