GEEscript_withOSMquery

// Supplement to the paper: "Breaking the ESG rating divergence: an open geospatial framework
// for environmental scores" by Rossi et al.

// Script compiled by Cristian Rossi and Jack Hepper.

// Steps you need to take: 

// STEP 1:
// Input a shapefile of the area to investigate. Click Assets (on left) -> New -> Upload. 
// Once processing has finished, import from Assets to script, and rename table to ald.

// STEP 2:
// Define the variables below. Input the start dates in 'YYYY-MM-DD' format. Input the
// monitoring period length of each period as a number of weeks. (4 is suggested.)

var firstPeriodStarts = '2020-08-01';
var secondPeriodStarts = '2022-08-01';
var monitoringPeriodWeeks = 4;

// STEP 3:
// Input variables on line 473 and 474 (Important!). 

// STEP 4 (FINAL): 
// Run script and use outputs under Console (on right) and Layers (below) to derive E-Score for
// asset. Note comments on lines 318, 375, and 451 to assist you in interpretation. An option to 
// generate an overpass turbo script for the OpenStreetMap analyses is also presented to the right 
// of the map, if you have run the script including OSM integration. Inspect CH4 and Dynamic World 
// (Classified Composite) layers closely; if there are too many NULL pixels (more than a few in
// Dynamic World and more than half of the AoI area for CH4), adjust monitoring periods until 
// layers draw satisfactorily.

// **************************************************************************************** //
// ************************** E Land Cover Change Score *********************************** //
// **************************************************************************************** //

// Script to compute newly anthropogenic areas from two land cover products 

// ---------------------------------------------------------------------------------------- //

// Display the shapefile into the interactive map

// Centre the view.

Map.centerObject(ald);

// Define styling and determine the color of the shapefile (hex starting
// with '00' denotes full transparency).

var styling = {color: 'red', fillColor: '00000000'};
Map.addLayer(ald.style(styling), '',  'AoI');

// --------------------------------------------------------------------------------------- //

// Dates
// First Monitoring Period
var startDate = firstPeriodStarts;
var endDate = ee.Date(startDate).advance(monitoringPeriodWeeks, 'week');
print('First monitoring period:', startDate, 'to', 
      (ee.Date(endDate).format()).slice(0,10)); 

// Second Monitoring Period
var startDate2 = secondPeriodStarts;
var endDate2 = ee.Date(startDate2).advance(monitoringPeriodWeeks, 'week');
print('Second monitoring period:', startDate2, 'to', 
      (ee.Date(endDate2).format()).slice(0,10));

// Creating multi-temporal composites from Dynamic World

// Filtering the Dynamic World NRT collection by monitoring period and location

var dwCollectionOne = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
  .filter(ee.Filter.date(startDate, endDate))
  .filter(ee.Filter.bounds(ald));

// Creating a mode composite from the temporally- and spatially-subset Dyanmic World collection

var classification = dwCollectionOne.select('label');
var dwComposite = classification.reduce(ee.Reducer.mode());

var dwVisParams = {
  min: 0,
  max: 8,
  palette: ['#419BDF', '#397D49', '#88B053', '#7A87C6',
    '#E49635', '#DFC35A', '#C4281B', '#A59B8F', '#B39FE1']
};

// Clipping the composite and adding it to the map

var title1 = ee.String('Classified Composite for Period Starting ')
    .cat(startDate.slice(0,4));
Map.addLayer(dwComposite.clip(ald), dwVisParams, title1.getInfo()); 

// Creating a Top-1 Probability Hillshade visualisation
// See developers.google.com/earth-engine/tutorials/community/introduction-to-dynamic-world-pt-1
// for explanation

var probabilityBands = [
    'water', 'trees', 'grass', 'flooded_vegetation', 'crops',
    'shrub_and_scrub', 'built', 'bare', 'snow_and_ice'
    ];

// Select probability bands 
var probabilityCol = dwCollectionOne.select(probabilityBands);

// Create a multi-band image with the average pixel-wise probability 
// for each band across the time-period
var meanProbability = probabilityCol.reduce(ee.Reducer.mean());

// Composites have a default projection that is not suitable
// for hillshade computation.
// Set a EPSG:3857 projection with 10m scale
var projection = ee.Projection('EPSG:3857').atScale(10);
var meanProbability = meanProbability.setDefaultProjection(projection);

// Creating the Top1 Probability Hillshade visualisation
var top1Probability = meanProbability.reduce(ee.Reducer.max());
var top1Confidence = top1Probability.multiply(100).int();
var hillshade = ee.Terrain.hillshade(top1Confidence).divide(255);
var rgbImage = dwComposite.visualize(dwVisParams).divide(255);
var probabilityHillshade = rgbImage.multiply(hillshade);

var hillshadeVisParams = {min:0, max:0.8};
var probabilityHillshadeTitle1 = ee.String('Probability Hillshade for Period Starting ')
    .cat(startDate.slice(0,4));
Map.addLayer(probabilityHillshade.clip(ald),
  hillshadeVisParams, probabilityHillshadeTitle1.getInfo());

// Creating multi-temporal composites from Dynamic World: Second period

// Filter the Dynamic World NRT collection by monitoring period and location

var dwCollectionTwo = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
  .filter(ee.Filter.date(startDate2, endDate2))
  .filter(ee.Filter.bounds(ald));

// Create a mode composite from the temporally- and spatially-subset Dyanmic World collection

var classification = dwCollectionTwo.select('label');
var dwComposite2 = classification.reduce(ee.Reducer.mode());

var dwVisParams = {
  min: 0,
  max: 8,
  palette: ['#419BDF', '#397D49', '#88B053', '#7A87C6',
    '#E49635', '#DFC35A', '#C4281B', '#A59B8F', '#B39FE1']
};

// Clip the composite and add it to the Map

var title2 = ee.String('Classified Composite for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(dwComposite2.clip(ald), dwVisParams, title2.getInfo());

// Create a Top-1 Probability Hillshade Visualization
var probabilityBands = [
    'water', 'trees', 'grass', 'flooded_vegetation', 'crops',
    'shrub_and_scrub', 'built', 'bare', 'snow_and_ice'
    ];

// Select probability bands 
var probabilityCol2 = dwCollectionTwo.select(probabilityBands);

// Create a multi-band image with the average pixel-wise probability 
// for each band across the time-period
var meanProbability2 = probabilityCol2.reduce(ee.Reducer.mean());

// Composites have a default projection that is not suitable
// for hillshade computation.
// Set a EPSG:3857 projection with 10m scale
var projection = ee.Projection('EPSG:3857').atScale(10);
var meanProbability2 = meanProbability2.setDefaultProjection(projection);

// Create the Top1 Probability Hillshade
var top1Probability2 = meanProbability2.reduce(ee.Reducer.max());
var top1Confidence2 = top1Probability2.multiply(100).int();
var hillshade2 = ee.Terrain.hillshade(top1Confidence2).divide(255);
var rgbImage2 = dwComposite2.visualize(dwVisParams).divide(255);
var probabilityHillshade2 = rgbImage2.multiply(hillshade2);

var hillshadeVisParams = {min:0, max:0.8};
var probabilityHillshadeTitle2 = ee.String('Probability Hillshade for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(probabilityHillshade2.clip(ald),
  hillshadeVisParams, probabilityHillshadeTitle2.getInfo());
  

// // ----------------------------------------------------------------------- //
// // Detect changes from natural to anthropogenic ------------------------- //

// anthropogenic
//built areas
var dw_built = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('built');

var beforeDw_built = dw_built.filterDate(startDate, endDate).mean();
var afterDw_built = dw_built.filterDate(startDate2, endDate2).mean();

// bare ground
var dw_bare = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('bare');

var beforeDw_bare = dw_bare.filterDate(startDate, endDate).mean();
var afterDw_bare = dw_bare.filterDate(startDate2, endDate2).mean();

// crops
var dw_crops = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('crops');

var beforeDw_crops = dw_crops.filterDate(startDate, endDate).mean();
var afterDw_crops = dw_crops.filterDate(startDate2, endDate2).mean();

// natural

//water
var dw_water = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('water');

var beforeDw_water = dw_water.filterDate(startDate, endDate).mean();
var afterDw_water = dw_water.filterDate(startDate2, endDate2).mean();

// trees
var dw_trees = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('trees');

var beforeDw_trees = dw_trees.filterDate(startDate, endDate).mean();
var afterDw_trees = dw_trees.filterDate(startDate2, endDate2).mean();

// grass	
var dw_grass = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('grass');

var beforeDw_grass = dw_grass.filterDate(startDate, endDate).mean();
var afterDw_grass	= dw_grass.filterDate(startDate2, endDate2).mean();

//flooded_vegetation	
var dw_flooded_vegetation = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('flooded_vegetation');

var beforeDw_flooded_vegetation = dw_flooded_vegetation.filterDate(startDate, endDate).mean();
var afterDw_flooded_vegetation = dw_flooded_vegetation.filterDate(startDate2, endDate2).mean();

// shrub_and_scrub
var dw_shrub_and_scrub = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('shrub_and_scrub');

var beforeDw_shrub_and_scrub = dw_shrub_and_scrub.filterDate(startDate, endDate).mean();
var afterDw_shrub_and_scrub = dw_shrub_and_scrub.filterDate(startDate2, endDate2).mean();

// snow and ice
var dw_snow_and_ice = ee.ImageCollection('GOOGLE/DYNAMICWORLD/V1')
            .filterBounds(ald)
            .select('snow_and_ice');

var beforeDw_snow_and_ice = dw_snow_and_ice.filterDate(startDate, endDate).mean();
var afterDw_snow_and_ice = dw_snow_and_ice.filterDate(startDate2, endDate2).mean();

// Selecting pixels that were water and are now anthropogenic

var newAntro_fromWater = beforeDw_water.gt(0.35).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.35)).or(afterDw_crops.gt(0.35)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro_fromWater.clip(ald), changeVisParams, 'New Anthropogenic from Water');

// Trees to Anthro

var newAntro_fromTrees = beforeDw_trees.gt(0.35).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.35)).or(afterDw_crops.gt(0.35)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro_fromTrees.clip(ald), changeVisParams, 'New Anthropogenic from Trees');

// Grass to Anthro

var newAntro_fromGrass = beforeDw_grass.gt(0.35).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.35)).or(afterDw_crops.gt(0.35)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro_fromGrass.clip(ald), changeVisParams, 'New Anthropogenic from Grass');

// Flooded Vegetation to Anthro

var newAntro_fromFlood = beforeDw_flooded_vegetation.gt(0.35).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.35)).or(afterDw_crops.gt(0.35)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro_fromFlood.clip(ald), changeVisParams, 'New Anthropogenic from Flooded Vegetation');

// Shrubs to Anthro

var newAntro_fromShrubs = beforeDw_shrub_and_scrub.gt(0.35).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.35)).or(afterDw_crops.gt(0.35)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro_fromShrubs.clip(ald), changeVisParams, 'New Anthropogenic from Shrubs');

// Display all new anthropogenic areas

var newAntro = (beforeDw_grass.gt(0.35).or(beforeDw_trees.gt(0.35)).or(beforeDw_water.gt(0.35)).or(beforeDw_flooded_vegetation.gt(0.35)).or(beforeDw_shrub_and_scrub.gt(0.35))).and((afterDw_built.gt(0.35)).or(afterDw_bare.gt(0.3)).or(afterDw_crops.gt(0.3)));

var changeVisParams = {min: 0, max: 1, palette: ['white', 'red']};
Map.addLayer(newAntro.clip(ald), changeVisParams, 'New Anthropogenic areas');
  

// ------------------------------------------------------------------------------ //
// ---- Visualisation check ------------------------------------------------------

var s2 = ee.ImageCollection('COPERNICUS/S2')
            .filterBounds(ald)
            .filter(ee.Filter.lt('CLOUDY_PIXEL_PERCENTAGE', 35));

var beforeS2 = s2.filterDate(startDate, endDate).median();
var afterS2 = s2.filterDate(startDate2, endDate2).median();

// Adjust stretch (min / max values) if images are too washed out; alternatively
// use settings accessible from layer menu on map

// Sentinel image one (earlier)

var s2VisParams1 = {bands: ['B4', 'B3', 'B2'], min: 0, max: 3000};
var s2VisTitle1 = ee.String('Sentinel-2 Image for Period Starting ')
    .cat(startDate.slice(0,4));
Map.addLayer(beforeS2.clip(ald), s2VisParams1, s2VisTitle1.getInfo());

// Sentinel image two (later)

var s2VisParams2 = {bands: ['B4', 'B3', 'B2'], min: 0, max: 3000};
var s2VisTitle2 = ee.String('Sentinel-2 Image for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(afterS2.clip(ald), s2VisParams2, s2VisTitle2.getInfo());

Map.addLayer(newAntro.selfMask().clip(ald), changeVisParams, 'New Anthropogenic areas (Masked)');
    
// ------------------------------------------------------------------------------ //    
// -------------- Find the percentage of change --------------------------------- //

var builtArea = dwComposite.eq(6).clip(ald);
var builtArea = builtArea.rename(['built_area']);

// Count all pixels.
var statsTotal = builtArea.reduceRegion({
    reducer: ee.Reducer.count(),
    geometry: ald,
    scale: 10,
    maxPixels: 1e10
    });
var totalPixels = statsTotal.get('built_area');

print('Total Pixels in AoI:', totalPixels);

// Mask 0 pixel values and count remaining pixels.
var builtAreaMasked = newAntro.selfMask().clip(ald);

var statsMasked = builtAreaMasked.reduceRegion({
    reducer: ee.Reducer.count(),
    geometry: ald,
    scale: 10,
    maxPixels: 1e10
    });
var builtAreaPixels = statsMasked.get('grass');   //takes the name of the first
print('Total pixels that changed from Natural to Anthropogenic in AoI:', builtAreaPixels);

var fraction = (ee.Number(builtAreaPixels).divide(totalPixels))
  .multiply(100);
print('Percentage of AoI that changed from Natural to Anthropogenic:', fraction.format('%.2f'));


// **************************************************************************************** //
// ***************************** E Hydrology Score **************************************** //
// **************************************************************************************** //

// Note: The most severe algal blooms may be misclassified by Dynamic World as flooded vegetation.
// Additionally, river islands and small land features adjacent to waterways can sometimes be detected
// as water. In both cases, this can result in pollution false negatives / positives. Be sure to 
// manually inspect the pollution hotspots and wider area in the Sentinel-2 visualisation layer to 
// double check validity of output. 

// 1. Extract Water Bodies from LCCs

// Extract the Water class.
var WaterArea1 = dwComposite.eq(0);
var WaterArea2 = dwComposite2.eq(0);

// 2. Mask Water areas

var WaterMasked1 = WaterArea1.clip(ald).updateMask(WaterArea1.eq(1));
var WaterMasked2 = WaterArea2.clip(ald).updateMask(WaterArea2.eq(1));

var waterMaskedTitle1 = ee.String('Water Areas for Period Starting ')
    .cat(startDate.slice(0,4));
Map.addLayer(WaterMasked1, {palette:['cyan']}, waterMaskedTitle1.getInfo());
var waterMaskedTitle2 = ee.String('Water Areas for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(WaterMasked2, {palette:['cyan']}, waterMaskedTitle2.getInfo());

// 3. Compute NDCI

var S2masked1 = beforeS2.updateMask(WaterArea1.eq(1));
var S2masked2 = afterS2.updateMask(WaterArea2.eq(1));

function ndci_f(img){
  //water mask
  var ndci = img.normalizedDifference(['B5', 'B4']).rename('NDCI');
  return img.addBands(ndci);
}

var before_NDCI = ndci_f(S2masked1);
var after_NDCI = ndci_f(S2masked2);


// print(IMG.get('CLOUDY_PIXEL_PERCENTAGE'))

var viz = {min:0.1,max:0.4,palette:['cyan','orange','red']};

var ndciTitle1 = ee.String('NDCI for Period Starting ')
    .cat(startDate.slice(0,4));
Map.addLayer(before_NDCI.select('NDCI'),viz,ndciTitle1.getInfo());
var ndciTitle2 = ee.String('NDCI for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(after_NDCI.select('NDCI'),viz,ndciTitle2.getInfo());


// Find high-values of NDCI

var before_NDCI_mask = before_NDCI.select('NDCI').gt(0.1);
var after_NDCI_mask = after_NDCI.select('NDCI').gt(0.1);
var before_NDCI_mask_title = ee.String('NDCI (Masked) for Period Starting ')
    .cat(startDate.slice(0,4));
var after_NDCI_mask_title = ee.String('NDCI (Masked) for Period Starting ')
    .cat(startDate2.slice(0,4));
Map.addLayer(before_NDCI_mask,{palette:['black','red']},before_NDCI_mask_title.getInfo());
Map.addLayer(after_NDCI_mask,{palette:['black','red']},after_NDCI_mask_title.getInfo());


// Overlay with Hydroshed to figure out drainage direction
var wwf = ee.Image('WWF/HydroSHEDS/03DIR');
var drainageDirection = wwf.select('b1').clip(ald);
var drainageDirectionVis = {
  min: 1.0,
  max: 128.0,
  palette: [
    '000000', '023858', '006837', '1a9850', '66bd63', 'a6d96a', 'd9ef8b',
    'ffffbf', 'fee08b', 'fdae61', 'f46d43', 'd73027'
  ],
};
Map.addLayer(drainageDirection, drainageDirectionVis, 'Drainage Direction (see line 451 for instruct)');

// Instructions for use:
// 1) Use the marker tool (📍) in the map view to define multiple points along a drainage basin.
// 2) Click the stop drawing tool (🤚).
// 3) Switch to the Inspector panel (on right).
// 4) Click the pushpins you have placed on the map to inspect their values. 
// To interpret the values, consult:
// developers.google.com/earth-engine/datasets/catalog/WWF_HydroSHEDS_03DIR#bands

//Clean water: NDCI values close to zero or slightly negative (around -0.1 to 0.1)
//Moderately polluted water: NDCI values between 0.1 and 0.4
//Highly polluted water: NDCI values above 0.4


// // **************************************************************************************** //
// // ************************** E Global Emissions Score ************************************ //
// // **************************************************************************************** //

// Important! Manually input dates to define dateRangeStart and dateRangeEnd, ensuring they do not fall 
// between 2022-07-26 and 2022-08-31, due to a provider outage during this period. The dates input 
// should correspond exactly to your second (later) monitoring period, or a close time period of
// equivalent length if your second monitoring period falls within the prohibited dates.

var dateRangeStart = '2022-09-01'; // add your number of weeks (default 4) to the date to the left to calculate the date below
var dateRangeEnd = '2022-10-01'; // check neither date falls within prohibited period

var collection = ee.ImageCollection('COPERNICUS/S5P/OFFL/L3_CH4')
  .select('CH4_column_volume_mixing_ratio_dry_air_bias_corrected')
  .filterDate(dateRangeStart, dateRangeEnd);

var band_viz = {
  min: 1800,
  max: 2000,
  palette: ['black', 'blue', 'purple', 'cyan', 'green', 'yellow', 'red']
};

var methaneSite = collection.mean().clip(ald);

Map.addLayer(methaneSite, band_viz, 'S5P CH4');


// Compute mean methane emission within site boundary
var meanMethane = methaneSite.reduceRegion({
                  reducer: ee.Reducer.mean(),
                  geometry: ald,
                  scale: 1000  //1km
                  });

// Print result to console
print('Mean methane emission over site:', meanMethane);  // low: <1800, mid: 1800-2000, high: >2000


///////////////////

var collection = ee.ImageCollection('COPERNICUS/S5P/OFFL/L3_SO2')
  .select('SO2_column_number_density')
  .filterDate(dateRangeStart, dateRangeEnd);

var band_viz = {
  min: 0.0,
  max: 0.001,
  palette: ['black', 'blue', 'purple', 'cyan', 'green', 'yellow', 'red']
};

var sulphurSite = collection.mean().clip(ald);
Map.addLayer(sulphurSite, band_viz, 'S5P SO2');

// Compute mean sulphur dioxide emission within site boundary
var meanSulphur = sulphurSite.reduceRegion({
                  reducer: ee.Reducer.mean(),
                  geometry: ald,
                  scale: 1000  //1km
                  });

// Print result to console
print('Mean sulfur dioxide emission over site:', meanSulphur);  // low: <0.001 mid:0.001-0.01 high:>0.1 


/////////

var collection = ee.ImageCollection('COPERNICUS/S5P/OFFL/L3_NO2')
  .select('tropospheric_NO2_column_number_density')
  .filterDate(dateRangeStart, dateRangeEnd);

var band_viz = {
  min: 0,
  max: 0.0002,
  palette: ['black', 'blue', 'purple', 'cyan', 'green', 'yellow', 'red']
};

var nitrogenSite = collection.mean().clip(ald);
Map.addLayer(nitrogenSite, band_viz, 'S5P N02');

// Compute mean nitrogen dioxide emission within site boundary
var meanNitrogen = nitrogenSite.reduceRegion({
                  reducer: ee.Reducer.mean(),
                  geometry: ald,
                  scale: 1000  //1km
                  });

// Print result to console
print('Mean nitrogen dioxide emission over site:', meanNitrogen);  // low: 0 - 0.0002 mid:0.0002 - 0.0005 high:>0.0005 



///////

var collection = ee.ImageCollection('COPERNICUS/S5P/OFFL/L3_CO')
  .select('CO_column_number_density')
  .filterDate(dateRangeStart, dateRangeEnd);

var band_viz = {
  min: 0,
  max: 0.05,
  palette: ['black', 'blue', 'purple', 'cyan', 'green', 'yellow', 'red']
};

var carbonSite = collection.mean().clip(ald);

Map.addLayer(carbonSite, band_viz, 'S5P CO');

// Compute mean carbon monoxide emission within site boundary
var meanCarbon = carbonSite.reduceRegion({
                  reducer: ee.Reducer.mean(),
                  geometry: ald,
                  scale: 1000  //1km
                  });

// Print result to console
print('Mean carbon monoxide emission over site:', meanCarbon);  // low:  0 - 0.01  mid:0.01 - 0.05 high:>0.05

// **************************************************************************************** //
// ******************************** OpenStreetMap ***************************************** //
// **************************************************************************************** //

// Extract AoI coordinates

var AoI = ee.FeatureCollection(ald).geometry();
var myCoordinates = AoI.coordinates();

// Convert coordinates to a client-side array

var coordinatesArray = myCoordinates.getInfo();

// Flip lon/lat pairs to lat/lon within each pair and extract as string

var numbersStringComma = coordinatesArray.map(function(coord) {
  var flippedPair = coord.map(function(num) {
    return [num[1], num[0]];
  });
  return flippedPair.join(',');
}).join(',');

// Duplicate and process string(s) for incorporation into URL paramaters

var poly = numbersStringComma.replace(/,/g, '%20');
var polyBoundary = numbersStringComma.replace(/,/g, '%2C');

// URL parameterisation

var urlFrag1 = "https://overpass-turbo.eu/?Q=%2F%2A+Supplement+to+the+paper%3A+%22Breaking+the+ESG+rating+divergence%3A+an+open+geospatial+framework+for+environmental+scores%22+by+Rossi+et+al.%0D%0A%0D%0AAoI+polygon+information+auto-imported+from+GEE+script.%0D%0A%0D%0AWhat+you+need+to+do%3A%0D%0A%0D%0ASTEP+1%3A+Define+variables+below+to+describe+what+to+check+AoI+for+and+how+far+to+buffer+around+it.%0D%0A%0D%0A%7B%7Btag%3Dprotected_area%7D%7D+Add+protected_area+%2F+nature_reserve+%2F+wetland%0D%0A%7B%7Bbuffer%3D0%7D%7D+Define+buffer+around+AoI+in+metres%0D%0A%0D%0ASTEP+2%3A+Click+%27Run%27.+Then+click+%27zoom+to+data%27+%28the+%F0%9F%94%8D+icon%29.%0D%0A%0D%0ASTEP+3+%28FINAL%29%3A+Inspect+output.+Areas+drawn+intersect+with+the+AoI+%28plus+buffer%2C+if+greater+than+zero%29.+If+nothing+is+drawn%2C+that+means+there+is+no+element+within+%28buffer+meters+of%29+the+AoI+with+the+tag+value.%0D%0A%0D%0A%7B%7Bpoly%3D";
var urlFrag2 = "%7D%7D%0D%0A%7B%7Bpolyboundary%3D";
var urlFrag3 = "%7D%7D%0D%0A%0D%0A%7B%7Bstyle%3A%0D%0A%09way%2C+relation%2C+node%0D%0A%09%7B+color%3Ared%3B+fill-color%3Awhite%3B+%7D%0D%0A%7D%7D%0D%0A%0D%0ANB+only+looks+for+ways%2Frelations+%28wr%29%2C+not+nodes+-+as+they+are+not+deemed+relevant+for+the+E-Score+analysis.+%2A%2F%0D%0A%0D%0Awr%5B%22boundary%22%3D%22%7B%7Btag%7D%7D%22%5D%28poly%3A%22%7B%7Bpoly%7D%7D%22%29%3B%28._%3B%3E%3B%29%3Bout+body%3B%0D%0A%0D%0Awr%5B%22boundary%22%3D%22%7B%7Btag%7D%7D%22%5D%28around%3A%7B%7Bbuffer%7D%7D%2C%7B%7Bpolyboundary%7D%7D%29%3B%28._%3B%3E%3B%29%3Bout+body%3B%0D%0A+++%0D%0Awr%5B%22leisure%22%3D%22%7B%7Btag%7D%7D%22%5D%28poly%3A%22%7B%7Bpoly%7D%7D%22%29%3B%28._%3B%3E%3B%29%3Bout+body%3B%0D%0A+++%0D%0Awr%5B%22leisure%22%3D%22%7B%7Btag%7D%7D%22%5D%28around%3A%7B%7Bbuffer%7D%7D%2C%7B%7Bpolyboundary%7D%7D%29%3B%28._%3B%3E%3B%29%3Bout+body%3B%0D%0A%0D%0Awr%5B%22natural%22%3D%22%7B%7Btag%7D%7D%22%5D%28poly%3A%22%7B%7Bpoly%7D%7D%22%29%3B%28._%3B%3E%3B%29%3Bout+body%3B%0D%0A+++%0D%0Awr%5B%22natural%22%3D%22%7B%7Btag%7D%7D%22%5D%28around%3A%7B%7Bbuffer%7D%7D%2C%7B%7Bpolyboundary%7D%7D%29%3B%28._%3B%3E%3B%29%3Bout+body%3B";

var url = urlFrag1 + poly + urlFrag2 + polyBoundary + urlFrag3;

// Add panel to right of map

var panel = ui.Panel({
  style: {
    width: '115px',
  }
});
ui.root.insert(1, panel);

// Add link to overpass turbo script

var link = ui.Label('Generate overpass turbo script for AoI', {
    fontWeight: 'bold',
    backgroundColor: 'fffb91' 
},url);
panel.add(link);

// Add warning text (can malform on MS Edge because of Edge's very low URL length 
// limit - as polygon info is passed via URL this causes issues)

var text1 = ui.Label('(For OpenStreetMap analyses)', {
  fontSize: '11.5px'
});
panel.add(text1);

var text2 = ui.Label('Please note: overpass turbo script generation assumes AoI is one solid polygon. If AoI is multiple polygons or a polygon with an internal void, script will produce incorrect output. Script can malform on Microsoft Edge; a different browser is recommended.', {
  fontSize: '10px'
});
panel.add(text2);