Thermal · Diurnal Temperature Range

LST_DIFF – Day–Night Land Surface Temperature Difference

LST_DIFF represents the diurnal land surface temperature difference, computed as daytime LST minus nighttime LST. It is useful for urban heat island analysis, surface thermal inertia, and drought and land–atmosphere interaction studies.

1. Definition

The LST_DIFF index quantifies how much hotter the land surface is during the day compared to the night.

Formula

LST_DIFF = LST_DAY − LST_NIGHT

  • LST_DAY – daytime land surface temperature (°C)
  • LST_NIGHT – nighttime land surface temperature (°C)

Interpretation

  • High positive LST_DIFF → strong diurnal cycle (surfaces heat up a lot during the day)
  • Low LST_DIFF → more thermally stable surfaces (water bodies, dense vegetation)

Applications

  • Urban Heat Island and urban canyon studies
  • Surface thermal inertia & material properties
  • Drought and land degradation assessment
  • Microclimate and climate change impact analysis

2. Example Dataset

MODIS (used in the code)

  • Product → MOD11A2.061 (Terra, 8-day, 1 km)
  • Bands → LST_Day_1km, LST_Night_1km
  • Scale factor → 0.02
  • Units → Kelvin (then converted to °C)

Suggested Palette (°C difference)

[ "#313695", "#4575b4", "#74add1", "#fdae61", "#f46d43", "#a50026" ]

3. Google Earth Engine Code – LST_DIFF (Day − Night, MODIS)


// LST_DIFF – Day–Night Land Surface Temperature Difference (°C)
// LST_DIFF = LST_DAY - LST_NIGHT

var roi = geometry;
Map.centerObject(roi, 6);

// 1. Load MODIS LST (Terra, 8-day, 1 km)
var lstCol = ee.ImageCollection("MODIS/061/MOD11A2")
  .filterBounds(roi)
  .filterDate("2023-01-01", "2023-12-31")
  .select(["LST_Day_1km", "LST_Night_1km"]);

// 2. Convert to Celsius for both day and night
// MODIS scale factor = 0.02, units Kelvin
function toCelsiusPair(img) {
  var dayK   = img.select("LST_Day_1km").multiply(0.02);
  var nightK = img.select("LST_Night_1km").multiply(0.02);
  
  var dayC   = dayK.subtract(273.15).rename("LST_DAY_C");
  var nightC = nightK.subtract(273.15).rename("LST_NIGHT_C");
  
  return dayC.addBands(nightC).copyProperties(img, img.propertyNames());
}

var lstPairCol = lstCol.map(toCelsiusPair);

// 3. Aggregate over time (e.g. mean of 2023)
var lstMean = lstPairCol.mean().clip(roi);

// 4. Compute LST_DIFF = LST_DAY_C - LST_NIGHT_C
var lstDiff = lstMean.select("LST_DAY_C")
  .subtract(lstMean.select("LST_NIGHT_C"))
  .rename("LST_DIFF_C");

// 5. Visualization (°C difference)
var vis = {
  min: 0,
  max: 25,
  palette: [
    "#313695", "#4575b4", "#74add1",
    "#fdae61", "#f46d43", "#a50026"
  ]
};

Map.addLayer(lstDiff, vis, "LST_DIFF (Day - Night, °C)");

// 6. Optional: high diurnal range mask (e.g. > 15 °C)
var highDiff = lstDiff.gt(15).selfMask();
Map.addLayer(highDiff, {palette:["#a50026"]}, "High Diurnal Range (>15°C)", false);

// 7. Export LST_DIFF
Export.image.toDrive({
  image: lstDiff,
  description: "LST_DIFF_export",
  fileNamePrefix: "LST_DIFF_MODIS_2023",
  region: roi,
  scale: 1000,   // MODIS ~1 km
  crs: "EPSG:4326",
  maxPixels: 1e13
});