GEE-code

## Google Earth Engine code for extracting enviromental variables for nightingale presence/absence records 

// Create bounding box for West African region
var wa = ee.Geometry.BBox(-18.0416195374957, 2.99999216327873, 15, 17.5);

// Zoom to region of West Africa
Map.setCenter(0, 10, 4); 
// Plot records
Map.addLayer(nigal, {}, 'Records');

// Bioclimate data 
var clim = ee.ImageCollection('IDAHO_EPSCOR/TERRACLIMATE')
    .filterDate('2001-01-01', '2020-12-31');

// ************************************ RAINFALL ********************************************** //                 

var prec = clim.select('pr');

// Look at precipiation in the region
// Set visualitation parameters
var prVis = { 
  min: 0,
  max: 150,
  palette: [
    '#fff7fb','#ece7f2','#d0d1e6','#a6bddb','#74a9cf','#3690c0','#0570b0','#045a8d','#023858'
    ],
};
// Plot map
Map.addLayer(prec, prVis, 'Accumulated rainfall');

// Extract annual rainfall. This will give an indication of whether the location experienced higher 
// or lower than average rainfall. Map through all records to match dates with images. 
var precRes = nigal.map(function (f) {
  
  var date = f.get('date'); // Date of record
  
  // Create new property specifying the year of interest
  var mn = ee.Number.parse(ee.String(date).slice(3,5));
  var yr = ee.Number.parse(ee.String(date).slice(6,10));
  // If the record is from Jan or Feb, we are interested in the year before
  var nyr = ee.Algorithms.If(mn.lt(3),yr.subtract(1),yr); 
  
  return ee.Feature(f.geometry(), 
  
  // Select images years of interest and sum values of images 
    prec.filter(ee.Filter.calendarRange(nyr,nyr,'year')).
         sum().
  
  // Extract total rainfall for each record
  reduceRegion({
    reducer: ee.Reducer.first(),
    geometry: f.geometry(), 
    scale: 4638.3,
  })).set('ID', f.get('ID')); // Retain ID variable
});

// Extract monthly rainfall
// Use a function to iterate through months and create images of average monthly precipitation
var months = ee.List.sequence(1,12); // Create list of months 
// Rather than doing this within the reduceRegions function, we want to export 
// images because we will need them for making predictions later on 
var prec_monthly = ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(prec.
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).toBands(). // Convert to multi-band image
    rename('JanRain','FebRain','MarchRain','AprRain','MayRain','JuneRain','JulyRain','AugRain','SepRain','OctRain','NovRain','DecRain');

print('Monthly precipitation:', prec_monthly.getInfo());

var monthlyPrecRes = prec_monthly.
  reduceRegions({
    reducer: ee.Reducer.first(),
    collection: nigal,
    scale: 4638.3,
  });

// Check results
print('Precipitation results:', precRes.first());
print(monthlyPrecRes.first());

// Export
Export.table.toDrive({
  collection: precRes,  
  description: 'nigal-prec',
  folder: 'nigal GEE results'
});
Export.table.toDrive({
  collection: monthlyPrecRes, 
  description: 'nigal-monthly-prec',
  folder: 'nigal GEE results'
});

// ************************************ TEMPERATURE ********************************************** //                 

var tmp = clim.select(['tmmn','tmmx']);

// Plot temp
var tmpVis = {
  min: -300.0,
  max: 300.0,
  palette: [
    '1a3678', '2955bc', '5699ff', '8dbae9', 'acd1ff', 'caebff', 'e5f9ff',
    'fdffb4', 'ffe6a2', 'ffc969', 'ffa12d', 'ff7c1f', 'ca531a', 'ff0000',
    'ab0000'
  ],
};
Map.addLayer(tmp.select('tmmn'), tmpVis, 'Temperature');

// Iterate through months and select average minimum and max temperature
var tmp_min = ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(tmp.select('tmmn').
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).min();
Map.addLayer(tmp_min, tmpVis, 'Minimum temperature');

var tmp_max = ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(tmp.select('tmmx').
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).max();

// Extract temp for each record
var maxTmpRes = tmp_max.
  reduceRegions({
    reducer: ee.Reducer.first().setOutputs(["tmmx"]), 
    collection: nigal,
    scale: 4638.3,
  });
  
var minTmpRes = tmp_min.
  reduceRegions({
    reducer: ee.Reducer.first().setOutputs(["tmmn"]), 
    collection: nigal,
    scale: 4638.3,
  });

// Check results
print('Temperature results:', minTmpRes.first());
print(maxTmpRes.first());

// Export
Export.table.toDrive({
  collection: minTmpRes, 
  description: 'nigal-min-temp',
  folder: 'nigal GEE results'
});

Export.table.toDrive({
  collection: maxTmpRes, 
  description: 'nigal-max-temp',
  folder: 'nigal GEE results'
});

// ************************************ PALMER'S DROUGHT SEVERITY INDEX ********************************************** //         

var pdsi = clim.select('pdsi');

// Visualise
var pdsiVis = {
  min: -4000.0,
  max: 3000.0,
  palette: [
    '1a3678', '2955bc', '5699ff', '8dbae9', 'acd1ff', 'caebff', 'e5f9ff',
    'fdffb4', 'ffe6a2', 'ffc969', 'ffa12d', 'ff7c1f', 'ca531a', 'ff0000',
    'ab0000'
  ],
};
Map.addLayer(pdsi, pdsiVis, 'PDSI');

// Repeat extraction of annual rainfall using PDSI
var pdsiRes = nigal.map(function (f) {
  
  // What year is the record from?
  var date = f.get('date'); 
  // If the record is from Jan or Feb, we are interested in the year before
  var mn = ee.Number.parse(ee.String(date).slice(3,5));
  var yr = ee.Number.parse(ee.String(date).slice(6,10));
  var nyr = ee.Algorithms.If(mn.lt(3),yr.subtract(1),yr); 
  
  return ee.Feature(f.geometry(), 
  
  // Select images years of interest and sum values of images 
    pdsi.filter(ee.Filter.calendarRange(nyr,nyr,'year')).
         mean().
  
  // Extract annual PDSI for each record
  reduceRegion({
    reducer: ee.Reducer.first(),
    geometry: f.geometry(), 
    scale: 4638.3,
  })).set('ID', f.get('ID')); // Retain ID variable;
});

// Check results
print('Drought results:', pdsiRes.first());

// Export
Export.table.toDrive({
  collection: pdsiRes, 
  description: 'nigal-pdsi',
  folder: 'nigal GEE results'
});

// ************************************ NDVI ********************************************** //                 

var terra = ee.ImageCollection('MODIS/061/MOD13Q1').select('NDVI');
var aqua = ee.ImageCollection('MODIS/061/MYD13Q1').select('NDVI');

// Merge image collections
var ndvi = terra.merge(aqua).
 filterDate('2001-01-01', '2020-12-31'); // This time period covers out study period and a bit before 
 
// Plot mean NDVI
var ndviVis = {
  min:0,
  max:9000,
  palette: ['ffffff','ce7e45','df923d','f1b555','fcd163','99b718','74a901','66a000','529400',
    '3e8601','207401','056201','004c00','023b01','012e01','011d01','011301'],
  }; 
Map.addLayer(ndvi, ndviVis, 'NDVI');

// Use a function to iterate through months and create images of average monthly NDVI. 
var ndvi_monthly = ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(ndvi.
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).toBands(). // Convert to multi-band image
    rename('JanNDVI','FebNDVI','MarchNDVI','AprNDVI','MayNDVI','JuneNDVI','JulyNDVI','AugNDVI','SepNDVI','OctNDVI','NovNDVI','DecNDVI');
// Explore monthly composites  
print('NDVI bands:', ndvi_monthly.getInfo());

// Extract data underlying nightingale points
var ndviRes = ndvi_monthly.reduceRegions({
  reducer: ee.Reducer.first(),
  collection: nigal,
  scale: 250,
  });

// Check results
print('NDVI results:', ndviRes.first());

// Export
Export.table.toDrive({
  collection: ndviRes, 
  description: 'nigal-ndvi',
  folder: 'nigal GEE results'
});

// ************************************ ELEVATION ********************************************** //                 

var elev = ee.Image('CGIAR/SRTM90_V4').select('elevation');
Map.addLayer(elev, {min: 0, max: 60}, 'Elevation');

var elevRes = elev.
  reduceRegions({
    reducer: ee.Reducer.first().
      setOutputs(['elevation']), 
    collection: nigal,
    scale: 92.76624232,
  });

// Check results
print('Elevation results:', elevRes.first());

// Export
Export.table.toDrive({
  collection: elevRes, 
  description: 'nigal-elev',
  folder: 'nigal GEE results'
});

// #### Export results #### //

// Combine properties in the different feature collections
// var res = precRes.map (function (f) {
//     return f.copyProperties(avgprecRes.filter(ee.Filter.eq('ID', f.get('ID'))).first()).
//              copyProperties(tmpRes.filter(ee.Filter.eq('ID', f.get('ID'))).first()).
//              copyProperties(pdsiRes.filter(ee.Filter.eq('ID', f.get('ID'))).first()).
//              copyProperties(ndviRes.filter(ee.Filter.eq('ID', f.get('ID'))).first()).
//              copyProperties(elevRes.filter(ee.Filter.eq('ID', f.get('ID'))).first());
//   });

// Check results 
// print('Final results:', res.first());

// Export results
// Export.table.toDrive({
//   collection: res
//   description: 'nigal-GEE-results',
//   folder: 'nigal GEE results'
// });

// Export rasters
var rastExp = prec_monthly.addBands(tmp_min).addBands(tmp_max).addBands(ndvi_monthly).addBands(pdsi.mean()).addBands(elev.double());
// Need to make elevation image into 64-bit float to match other images
print('Combined rasters:', rastExp); 

Export.image.toDrive({ 
  image: rastExp,
  description: 'GEE-rasts',
  // Crop by West African extent
  region: wa,
  folder: 'nigal GEE results',
  maxPixels: 1e12,
  scale: 4638 // Match lowest resolution of climate data
});

// Export rasters to look at change over time 
var rast2020 = ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(prec
        .filterDate('2020-01-01', '2020-12-31')
        .filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).toBands(). // Convert to multi-band image
    rename('JanRain','FebRain','MarchRain','AprRain','MayRain','JuneRain','JulyRain','AugRain','SepRain','OctRain','NovRain','DecRain')
  .addBands(ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(ndvi
        .filterDate('2020-01-01', '2020-12-31')
        .filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).toBands(). // Convert to multi-band image
    rename('JanNDVI','FebNDVI','MarchNDVI','AprNDVI','MayNDVI','JuneNDVI','JulyNDVI','AugNDVI','SepNDVI','OctNDVI','NovNDVI','DecNDVI'))
    .addBands(ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(tmp.select('tmmx').filterDate('2020-01-01', '2020-12-31').
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).max())
    .addBands(ee.ImageCollection.fromImages(
      months.map(function (m) {
        return ee.Image(tmp.select('tmmn').filterDate('2020-01-01', '2020-12-31').
        filter(ee.Filter.calendarRange(m, m, 'month'))
                    .mean());
})).min())
  .addBands(pdsi.filterDate('2020-01-01', '2020-12-31').mean())
  .addBands(elev.double());

Export.image.toDrive({ 
  image: rast2020,
  description: '2020-rasts',
  // Crop by West African extent
  region: wa,
  folder: 'nigal GEE results',
  maxPixels: 1e12,
  scale: 4638 // Match lowest resolution of climate data
});