Debugging Guide for Earth Engine

The debugging mindset

1. Read the error carefully. The message often tells you exactly what's wrong.
2. Isolate the problem. Comment out code until it works, then add back line by line.
3. Print intermediate results. Use print() to inspect objects.
4. Simplify. Test on a tiny area or single image first.
5. Search. Others have likely had the same problem.

Common errors and fixes

1. "X is not a function"

Cause: The variable isn't what you think it is (maybe null or wrong type).

Fix:

// Debug: Check what the variable actually is
print('Type of image:', image);

// Common cause: .first() on empty collection
var collection = ee.ImageCollection('...')
  .filterDate('2023-01-01', '2023-01-02')
  .filterBounds(roi);
print('Collection size:', collection.size()); // Check if empty!
var image = collection.first();

2. "Image.select: Pattern 'X' did not match any bands"

Cause: Wrong band name for this dataset.

Fix:

// Debug: Print actual band names
print('Available bands:', image.bandNames());

// Landsat 8 SR uses SR_B5, not B5
var nir = image.select('SR_B5'); // Correct for Landsat 8 SR

3. "User memory limit exceeded"

Cause: Too much data being processed at once.

Fix:

// 1. Reduce the area
var smallRoi = roi.buffer(-5000); // Shrink your ROI

// 2. Increase scale (lower resolution)
var result = image.reduceRegion({
  reducer: ee.Reducer.mean(),
  geometry: roi,
  scale: 100,  // Was 30, now 100
  maxPixels: 1e9
});

// 3. Use bestEffort
var result = image.reduceRegion({
  reducer: ee.Reducer.mean(),
  geometry: roi,
  scale: 30,
  bestEffort: true  // Auto-adjust scale if needed
});

4. "Cannot add an ee.Image to a number"

Cause: Mixing client and server objects with JavaScript operators.

Fix:

// Wrong: Using + with server objects
var bad = image.add(10 + ee.Number(5));

// Right: Use ee methods
var good = image.add(ee.Number(10).add(5));

5. "Computation timed out"

Cause: The computation is too complex or the area too large.

Fix:

// 1. Export instead of print for large results
Export.image.toDrive({...});

// 2. Reduce collection size
var filtered = collection
  .filterDate('2023-06-01', '2023-06-30')  // Shorter date range
  .filter(ee.Filter.lt('CLOUD_COVER', 20)); // Fewer images

// 3. Simplify the computation
// Instead of 100 bands, select only what you need
var simple = image.select(['SR_B4', 'SR_B5']);

6. "Dictionary does not contain key: X"

Cause: The property or key you're accessing doesn't exist.

Fix:

// Debug: Print all keys
var result = image.reduceRegion({...});
print('Keys:', result.keys());

// Make sure you're using the correct property name
var mean = result.get('SR_B5'); // Use actual key name

7. Empty or null results

Cause: Filters removed all data, or the geometry doesn't overlap.

Fix:

// Check collection size at each step
var step1 = ee.ImageCollection('LANDSAT/LC08/C02/T1_L2');
print('All images:', step1.size());

var step2 = step1.filterDate('2023-06-01', '2023-08-31');
print('After date filter:', step2.size());

var step3 = step2.filterBounds(roi);
print('After bounds filter:', step3.size());

// If 0, check your ROI location and date range

Client vs. Server: Quick reference

Rule: Once you're in server-land (ee. objects), stay there. Don't try to use JavaScript operators on Earth Engine objects.

Debug with print()

Strategic print statements are your best friend:

// Print at each step to find where things go wrong
var collection = ee.ImageCollection('LANDSAT/LC08/C02/T1_L2')
  .filterDate('2023-06-01', '2023-08-31')
  .filterBounds(roi);
print('1. Collection size:', collection.size());
print('2. First image:', collection.first());
print('3. Band names:', collection.first().bandNames());

var ndvi = collection.map(function(img) {
  return img.normalizedDifference(['SR_B5', 'SR_B4']);
});
print('4. NDVI collection:', ndvi.first());

var composite = ndvi.median();
print('5. Composite:', composite);

Scale matters!

Earth Engine resamples everything to match your specified scale. Wrong scale = wrong results.

// Always specify scale in reduceRegion
var stats = image.reduceRegion({
  reducer: ee.Reducer.mean(),
  geometry: roi,
  scale: 30,  // Match your data's native resolution!
  maxPixels: 1e9
});

// Common native resolutions:
// Landsat: 30m
// Sentinel-2: 10m (visible), 20m (red edge), 60m (atmospheric)
// MODIS: 250m, 500m, 1000m
// DEM (SRTM): 30m

When to ask for help

Before posting a question:

1. ✅ Searched for the error message
2. ✅ Read the function documentation
3. ✅ Created a minimal example that reproduces the error
4. ✅ Included the full error message
5. ✅ Described what you expected vs. what happened

Related resources

• Client vs. Server Objects
• Official GEE Debugging Guide
• Best Practices Guide