What You Need
Before you begin, ensure you have access to the following tools and data sources:
- Satellite imagery platforms – such as NASA Earth Observatory, USGS EarthExplorer, or ESA Copernicus Open Access Hub for Landsat 8/9 or Sentinel-2 images.
- Precipitation data – from NASA’s Global Precipitation Measurement (GPM) mission or IMERG to estimate rainfall totals.
- Topographic maps or digital elevation models – to identify steep slopes and drainage basins.
- Cyclone track and intensity records – from meteorological agencies like the Australian Bureau of Meteorology or the Joint Typhoon Warning Center.
- Basic image analysis skills – ability to compare two satellite scenes and recognize changes in surface cover.
- Reference images – before-and-after pairs of the same area, ideally acquired under cloud-free conditions.
Step-by-Step Guide
Step 1: Monitor Tropical Cyclone Conditions
Start by tracking the formation and movement of tropical cyclones in regions prone to landslides. Pay attention to areas near the equator (e.g., Papua New Guinea) where cyclones are less frequent but can be devastating. Look for unusually warm sea surface temperatures (above 26.5°C) and favorable atmospheric conditions that allow storms to intensify. For instance, Tropical Cyclone Maila reached Category 4 strength (on Australia’s scale) in April 2026 because of such conditions. Slow-moving storms are especially dangerous as they drench the same area repeatedly.
Step 2: Identify High-Risk Terrain
Focus on mountainous or hilly regions with steep slopes, especially where deforestation or previous landslides have exposed soil. Using digital elevation models, mark areas with slopes greater than 15–20 degrees. In the Gazelle district of East New Britain, the Baining Mountains are a classic example. The terrain’s steepness combined with heavy rain makes it vulnerable to slope failure.
Step 3: Gather Pre-cyclone Satellite Imagery
Acquire cloud-free satellite images of the target area taken before the cyclone’s rains began. For the Papua New Guinea example, a Landsat 9 image from September 24, 2025 shows intact dense green forest and clear waterways. This image serves as your baseline to detect changes. Ensure the image has a resolution fine enough (15–30 m) to spot landslide scars.
Step 4: Collect Post-cyclone Imagery
Obtain images captured soon after the cyclone passed and the rain stopped. Ideally, wait for a break in cloud cover. In the case of Cyclone Maila, a Landsat 9 image from April 20, 2026 – about 11 days after the landslides – revealed fresh scars. Use the same sensor and similar acquisition time of day to minimize differences in lighting.
Step 5: Compare Images to Detect Landslides
Place the before-and-after images side by side (or use a swipe tool). Look for light-brown swaths of exposed soil and debris cutting through darker green vegetation. These are fresh landslide paths. In the April 2026 image, such features appear running north toward a river valley. Note the contrast with surrounding green canopy and scattered white clouds. The shape is typically elongated, often with a fan-like deposit at the bottom.
Step 6: Examine Waterways for Sediment
Check rivers and streams in the post-cyclone image. Sediment-laden water appears light brown or milky compared to clear water in the pre-cyclone image. The Toriu River and other channels east of the landslides in the 2026 image show this discoloration. Increased sediment indicates that debris from landslides entered the drainage system.
Step 7: Correlate with Rainfall Estimates
Use satellite-based precipitation data (e.g., from GPM) to confirm that the area received extreme rainfall over a short period. In the Maila case, rainbands repeatedly hit East New Britain because the storm lingered. Compare totals to local thresholds for landslide initiation – often 100–200 mm in 24 hours for steep tropical slopes. This step validates that the observed scars were indeed rain-triggered.
Step 8: Verify with Ground Reports
Cross-reference your findings with local news reports, social media, or official disaster bulletins. For the Papua New Guinea event, news articles reported that landslides led to several deaths on and around April 9, 2026. Ground truth ensures that the features you see are indeed landslides and not logging, mining, or agricultural clearings.
Tips for Success
- Use multiple satellite sensors – Landsat, Sentinel-2, and MODIS can provide different resolutions and revisit times to increase your chances of cloud-free views.
- Account for cloud cover – In tropical regions, clouds often persist. Look for images taken during brief gaps, or use radar satellites (e.g., Sentinel-1) that can see through clouds.
- Watch for false positives – Shadows, new agriculture, or seasonal changes can mimic landslide scars. Compare with historical imagery from multiple years.
- Consider the Coriolis effect – Near the equator, cyclones are rare, but when they occur (as with Maila), the risk of landslides increases dramatically because the population is less prepared.
- Share your findings – If you identify unreported landslides, notify local disaster management agencies promptly. Your analysis could save lives.