Hands-on: Reading the landscape with SRTM terrain data

Terrain influences forest composition, soil moisture, accessibility, disturbance patterns, and harvesting operations. In this activity, you will use SRTM elevation data to create basic terrain products in GEE.

This exercise introduces simple raster operations, including loading an image, clipping it to a region of interest (ROI), calculating terrain derivatives, displaying map layers, summarizing raster values, and exporting a result.

Learning objectives¶

By the end of this activity, you should be able to:

• load a raster dataset in GEE

• define and display an ROI

• clip a raster to a study area

• derive slope, aspect, and hillshade from elevation

• calculate summary statistics for a raster layer

• export a raster product to Google Drive

Concept¶

The Shuttle Radar Topography Mission (SRTM) provides elevation data. Elevation is useful on its own, but it can also be used to derive other terrain variables that help describe the landscape.

In this activity, we create three common terrain products:

• Slope: the steepness of the surface

• Aspect: the direction a slope faces

• Hillshade: a shaded relief layer used for visual interpretation

These layers can help explain why forest conditions, species composition, disturbance patterns, and management activities vary across a landscape.

Complete GEE script¶

Copy the script below into the Google Earth Engine Code Editor.

// ------------------------------------------------------------
// SRTM terrain products for the Forêt Duparquet ROI
// ------------------------------------------------------------

// 1. Define the region of interest
var roi = ee.Geometry.Polygon([
  [
    [-79.50241092370698, 48.3829536766289],
    [-79.1741943709726, 48.3829536766289],
    [-79.1741943709726, 48.53958791639697],
    [-79.50241092370698, 48.53958791639697],
    [-79.50241092370698, 48.3829536766289]
  ]
]);

// 2. Load SRTM 30 m elevation data
var srtm = ee.Image("USGS/SRTMGL1_003");

// 3. Clip elevation to the ROI
var dem = srtm.clip(roi);

// 4. Derive terrain products
var slope = ee.Terrain.slope(dem);
var aspect = ee.Terrain.aspect(dem);
var hillshade = ee.Terrain.hillshade(dem);

// 5. Visualization parameters
var demVis = {
  min: 250,
  max: 450,
  palette: [
    "#006400",
    "#7FFF00",
    "#FFFF00",
    "#FFA500",
    "#8B4513"
  ]
};

var slopeVis = {
  min: 0,
  max: 30,
  palette: ["white", "yellow", "orange", "red"]
};

var aspectVis = {
  min: 0,
  max: 360,
  palette: [
    "blue", "cyan", "green", "yellow", "orange", "red", "purple"
  ]
};

var hillshadeVis = {
  min: 0,
  max: 255,
  palette: ["black", "white"]
};

// 6. Display layers
Map.centerObject(roi, 11);

Map.addLayer(dem, demVis, "SRTM elevation");
Map.addLayer(slope, slopeVis, "Slope");
Map.addLayer(aspect, aspectVis, "Aspect");
Map.addLayer(hillshade, hillshadeVis, "Hillshade");
Map.addLayer(roi, {color: "red"}, "ROI", false);

// 7. Calculate elevation statistics
var elevStats = dem.reduceRegion({
  reducer: ee.Reducer.minMax()
    .combine({
      reducer2: ee.Reducer.mean(),
      sharedInputs: true
    })
    .combine({
      reducer2: ee.Reducer.stdDev(),
      sharedInputs: true
    }),
  geometry: roi,
  scale: 30,
  maxPixels: 1e13
});

print("Elevation statistics for the ROI:", elevStats);

// ------------------------------------------------------------
// Export DEM to Google Drive
// ------------------------------------------------------------

// Export.image.toDrive({
//   image: dem,
//   description: "srtm_elevation_forest_duparquet",
//   folder: "GEE_exports",
//   fileNamePrefix: "srtm_elevation_forest_duparquet",
//   region: roi,
//   scale: 30,
//   crs: "EPSG:32617",
//   maxPixels: 1e13
// });

Script explanation¶

Define the ROI¶

The ROI is a rectangular study area around Forêt Duparquet, not the exact forest boundary. It is used to limit the analysis to the area of interest.

var roi = ee.Geometry.Polygon([...]);

The coordinates define the boundary of the study area. All raster operations in this exercise are clipped or summarized using this polygon.

Load SRTM¶

var srtm = ee.Image("USGS/SRTMGL1_003");

This line loads the SRTM 30 m elevation dataset from the Earth Engine data catalog. The dataset is loaded as an ee.Image, which is Earth Engine’s object type for raster data.

Clip the DEM¶

var dem = srtm.clip(roi);

Clipping keeps only the portion of the elevation raster inside the ROI. This does not change the original dataset. It creates a new image for display, analysis, and export.

Create terrain products¶

var slope = ee.Terrain.slope(dem);
var aspect = ee.Terrain.aspect(dem);
var hillshade = ee.Terrain.hillshade(dem);

Earth Engine provides built-in terrain functions. These functions use the elevation image to calculate slope, aspect, and hillshade.

Slope is measured in degrees from 0 to 90, where 0 is flat and larger values indicate steeper terrain.

Aspect is measured in degrees from 0 to 360 and represents the direction a slope faces: north is 0 or 360, east is 90, south is 180, and west is 270.

Hillshade is a visual layer that simulates terrain illumination, which makes landforms easier to interpret. In Earth Engine, the default hillshade uses light from the west, with an azimuth of 270° and an elevation angle of 45° above the horizon. It is used for visual interpretation, not as a physical measure of sunlight.

Display layers¶

Map.addLayer(dem, demVis, "SRTM elevation");
Map.addLayer(slope, slopeVis, "Slope");
Map.addLayer(aspect, aspectVis, "Aspect");
Map.addLayer(hillshade, hillshadeVis, "Hillshade");

Map.addLayer() displays raster or vector data in the Earth Engine Code Editor map. The visualization parameters control the color palette and value range used for display.

Calculate elevation statistics¶

var elevStats = dem.reduceRegion({
  reducer: ee.Reducer.minMax()
    .combine({
      reducer2: ee.Reducer.mean(),
      sharedInputs: true
    })
    .combine({
      reducer2: ee.Reducer.stdDev(),
      sharedInputs: true
    }),
  geometry: roi,
  scale: 30,
  maxPixels: 1e13
});

reduceRegion() summarizes raster values inside a geometry. Here, it calculates minimum, maximum, mean, and standard deviation of elevation within the ROI.

Export the DEM¶

Export.image.toDrive({
  image: dem,
  description: "srtm_elevation_forest_duparquet",
  folder: "GEE_exports",
  fileNamePrefix: "srtm_elevation_forest_duparquet", // if Earth Engine splits the export into tiles
  region: roi,
  scale: 30,
  crs: "EPSG:32617",
  maxPixels: 1e13
});

This block creates an export task for the clipped DEM in the Earth Engine Tasks tab. After running the script, open the Tasks tab and click Run to save the raster to Google Drive.

The export uses EPSG:32617, which is WGS 84 / UTM Zone 17N. This CRS is suitable because the ROI falls within UTM Zone 17. If you need to align the raster with Québec provincial datasets, EPSG:32198 can also be used. This corresponds to NAD83 / Québec Lambert, a common projection for Québec-wide geospatial data.

Expected outputs¶

After running the script, you should see:

• the ROI boundary on the map

• an elevation layer clipped to the study area

• a slope layer

• an aspect layer

• a hillshade layer

• elevation summary statistics printed in the Console

If you uncomment the export block, an export task will appear in the Tasks tab of the Earth Engine Code Editor.

Result example¶

Elevation, slope, aspect, and hillshade derived from SRTM for the rectangular study area around Forêt Duparquet.

Interpretation notes¶

Higher elevations may indicate ridges or upland areas. Lower elevations may correspond to wetlands, lakeshores, or low-lying terrain.

Steeper slopes may influence accessibility, road placement, harvesting operations, soil drainage, erosion risk, and local forest structure.

Aspect can influence incoming solar radiation, snowmelt, soil moisture, and vegetation composition. In northern landscapes, south-facing and north-facing slopes may differ in microclimate and growing conditions.

Hillshade is mainly a visualization product. It helps interpret landforms but should not be used as a quantitative terrain variable.

Questions¶

1. Where are the highest elevations in the ROI?

2. Where are the steepest slopes?

3. What landforms are visible in the terrain layers?

4. How could terrain affect forest access or management operations?

5. Which terrain product is easiest to interpret visually?

6. Which terrain variables could be useful predictors in a forest classification model?

Optional extension¶

Try changing the visualization ranges.

For example, modify the elevation range:

var demVis = {
  min: 270,
  max: 420,
  palette: ["green", "yellow", "brown"]
};

Or change the reducer used to calculate elevation statistics. For example, replace ee.Reducer.minMax() with:

ee.Reducer.median()   // median value

Then rerun the script and compare the map. Visualization settings do not change the data values, but they strongly affect how patterns appear on the map.