Here is my code.

var roi =table var s1 =  ee.ImageCollection('COPERNICUS/S1_GRD') .filter(ee.Filter.listContains('transmitterReceiverPolarisation', 'VH')) .filter(ee.Filter.listContains('transmitterReceiverPolarisation', 'VV')) //.filter(ee.Filter.eq('orbitProperties_pass', ['ASCENDING', 'DESCENDING'])) .filter(ee.Filter.eq('instrumentMode', 'IW')) .filterBounds(roi) .filterDate("2018-07-8","2018-08-26"); var firstNoTerrainCorrection = ee.Image(s1.median()).clip(roi); //Map.addLayer(firstNoTerrainCorrection,{min:-25,max:20},"no terrain correction"); Map.centerObject(roi, 5); s1 = s1.map(terrainCorrection); var s1_lee_sigma = s1.map(refinedLee); var firstTerrainCorrection = ee.Image(s1.first()); var s1_refinedLee  = ee.Image(s1_lee_sigma.median()); //print('s1_refinedLee',s1_refinedLee); s1_refinedLee = s1_refinedLee.clip(roi); var S1TerrainCorrection = firstTerrainCorrection.clip(roi); //Map.addLayer(s1_lee_sigma,{min:-25,max:20},"s1_lee_sigma") Map.addLayer(s1_refinedLee,{min:-25,max:20},"refined lee"); //Map.addLayer(s1,{min:-25,max:20},"s1") // Implementation by Andreas Vollrath (ESA), inspired by Johannes Reiche (Wageningen) function Clip(image,roi){   var mask=ee.Image.constant(1).clip(roi).mask();   return image.updateMask(mask); } function terrainCorrection(image) {   var imgGeom = image.geometry();   var srtm = ee.Image('USGS/SRTMGL1_003').clip(imgGeom); // 30m srtm   var sigma0Pow = ee.Image.constant(10).pow(image.divide(10.0));   // Article ( numbers relate to chapters)   // 2.1.1 Radar geometry   var theta_i = image.select('angle');   var phi_i = ee.Terrain.aspect(theta_i)     .reduceRegion(ee.Reducer.mean(), theta_i.get('system:footprint'), 1000)     .get('aspect');   // 2.1.2 Terrain geometry   var alpha_s = ee.Terrain.slope(srtm).select('slope');   var phi_s = ee.Terrain.aspect(srtm).select('aspect');   // 2.1.3 Model geometry   // reduce to 3 angle   var phi_r = ee.Image.constant(phi_i).subtract(phi_s);   // convert all to radians   var phi_rRad = phi_r.multiply(Math.PI / 180);   var alpha_sRad = alpha_s.multiply(Math.PI / 180);   var theta_iRad = theta_i.multiply(Math.PI / 180);   var ninetyRad = ee.Image.constant(90).multiply(Math.PI / 180);   // slope steepness in range (eq. 2)   var alpha_r = (alpha_sRad.tan().multiply(phi_rRad.cos())).atan();   // slope steepness in azimuth (eq 3)   var alpha_az = (alpha_sRad.tan().multiply(phi_rRad.sin())).atan();   // local incidence angle (eq. 4)   var theta_lia = (alpha_az.cos().multiply((theta_iRad.subtract(alpha_r)).cos())).acos();   var theta_liaDeg = theta_lia.multiply(180 / Math.PI);   // 2.2   // Gamma_nought_flat   var gamma0 = sigma0Pow.divide(theta_iRad.cos());   var gamma0dB = ee.Image.constant(10).multiply(gamma0.log10());   var ratio_1 = gamma0dB.select('VV').subtract(gamma0dB.select('VH'));   // Volumetric Model   var nominator = (ninetyRad.subtract(theta_iRad).add(alpha_r)).tan();   var denominator = (ninetyRad.subtract(theta_iRad)).tan();   var volModel = (nominator.divide(denominator)).abs();   // apply model   var gamma0_Volume = gamma0.divide(volModel);   var gamma0_VolumeDB = ee.Image.constant(10).multiply(gamma0_Volume.log10());   // we add a layover/shadow maskto the original implmentation   // layover, where slope > radar viewing angle   var alpha_rDeg = alpha_r.multiply(180 / Math.PI);   var layover = alpha_rDeg.lt(theta_i);   // shadow where LIA > 90   var shadow = theta_liaDeg.lt(90);   // calculate the ratio for RGB vis   var ratio = gamma0_VolumeDB.select('VV').subtract(gamma0_VolumeDB.select('VH'));   var output = gamma0_VolumeDB.addBands(ratio).addBands(alpha_r).addBands(phi_s).addBands(theta_iRad)     .addBands(layover).addBands(shadow).addBands(gamma0dB).addBands(ratio_1);   return image.addBands(     output.select(['VV', 'VH'], ['VV', 'VH']),     null,     true   ); } function powerToDb(img){   return ee.Image(10).multiply(img.log10()); } function dbToPower(img){   return ee.Image(10).pow(img.divide(10)); } // The RL speckle filter function refinedLee(image) {   var bandNames = image.bandNames();   image = dbToPower(image);   var result = ee.ImageCollection(bandNames.map(function(b){     var img = image.select([b]);     // img must be in natural units, i.e. not in dB!     // Set up 3x3 kernels     var weights3 = ee.List.repeat(ee.List.repeat(1,3),3);     var kernel3 = ee.Kernel.fixed(3,3, weights3, 1, 1, false);     var mean3 = img.reduceNeighborhood(ee.Reducer.mean(), kernel3);     var variance3 = img.reduceNeighborhood(ee.Reducer.variance(), kernel3);     // Use a sample of the 3x3 windows inside a 7x7 windows to determine gradients and directions     var sample_weights = ee.List([[0,0,0,0,0,0,0], [0,1,0,1,0,1,0],[0,0,0,0,0,0,0], [0,1,0,1,0,1,0], [0,0,0,0,0,0,0], [0,1,0,1,0,1,0],[0,0,0,0,0,0,0]]);     var sample_kernel = ee.Kernel.fixed(7,7, sample_weights, 3,3, false);     // Calculate mean and variance for the sampled windows and store as 9 bands     var sample_mean = mean3.neighborhoodToBands(sample_kernel);     var sample_var = variance3.neighborhoodToBands(sample_kernel);     // Determine the 4 gradients for the sampled windows     var gradients = sample_mean.select(1).subtract(sample_mean.select(7)).abs();     gradients = gradients.addBands(sample_mean.select(6).subtract(sample_mean.select(2)).abs());     gradients = gradients.addBands(sample_mean.select(3).subtract(sample_mean.select(5)).abs());     gradients = gradients.addBands(sample_mean.select(0).subtract(sample_mean.select(8)).abs());     // And find the maximum gradient amongst gradient bands     var max_gradient = gradients.reduce(ee.Reducer.max());     // Create a mask for band pixels that are the maximum gradient     var gradmask = gradients.eq(max_gradient);     // duplicate gradmask bands: each gradient represents 2 directions     gradmask = gradmask.addBands(gradmask);     // Determine the 8 directions     var directions = sample_mean.select(1).subtract(sample_mean.select(4))                                 .gt(sample_mean.select(4).subtract(sample_mean.select(7)))                                 .multiply(1);     directions = directions.addBands(sample_mean.select(6).subtract(sample_mean.select(4))                            .gt(sample_mean.select(4).subtract(sample_mean.select(2)))                            .multiply(2));     directions = directions.addBands(sample_mean.select(3).subtract(sample_mean.select(4))                            .gt(sample_mean.select(4).subtract(sample_mean.select(5)))                            .multiply(3));     directions = directions.addBands(sample_mean.select(0).subtract(sample_mean.select(4))                            .gt(sample_mean.select(4).subtract(sample_mean.select(8)))                            .multiply(4));     // The next 4 are the not() of the previous 4     directions = directions.addBands(directions.select(0).not().multiply(5));     directions = directions.addBands(directions.select(1).not().multiply(6));     directions = directions.addBands(directions.select(2).not().multiply(7));     directions = directions.addBands(directions.select(3).not().multiply(8));     // Mask all values that are not 1-8     directions = directions.updateMask(gradmask);     // "collapse" the stack into a singe band image (due to masking, each pixel has just one value (1-8) in it's directional band, and is otherwise masked)     directions = directions.reduce(ee.Reducer.sum());       //var pal = ['ffffff','ff0000','ffff00', '00ff00', '00ffff', '0000ff', 'ff00ff', '000000'];     //Map.addLayer(directions.reduce(ee.Reducer.sum()), {min:1, max:8, palette: pal}, 'Directions', false);     var sample_stats = sample_var.divide(sample_mean.multiply(sample_mean));     // Calculate localNoiseVariance     var sigmaV = sample_stats.toArray().arraySort().arraySlice(0,0,5).arrayReduce(ee.Reducer.mean(), [0]);     // Set up the 7*7 kernels for directional statistics     var rect_weights = ee.List.repeat(ee.List.repeat(0,7),3).cat(ee.List.repeat(ee.List.repeat(1,7),4));     var diag_weights = ee.List([[1,0,0,0,0,0,0], [1,1,0,0,0,0,0], [1,1,1,0,0,0,0],       [1,1,1,1,0,0,0], [1,1,1,1,1,0,0], [1,1,1,1,1,1,0], [1,1,1,1,1,1,1]]);     var rect_kernel = ee.Kernel.fixed(7,7, rect_weights, 3, 3, false);     var diag_kernel = ee.Kernel.fixed(7,7, diag_weights, 3, 3, false);     // Create stacks for mean and variance using the original kernels. Mask with relevant direction.     var dir_mean = img.reduceNeighborhood(ee.Reducer.mean(), rect_kernel)//.updateMask(directions.eq(1));     var dir_var = img.reduceNeighborhood(ee.Reducer.variance(), rect_kernel)//.updateMask(directions.eq(1));     dir_mean = dir_mean.addBands(img.reduceNeighborhood(ee.Reducer.mean(), diag_kernel).updateMask(directions.eq(2)));     dir_var = dir_var.addBands(img.reduceNeighborhood(ee.Reducer.variance(), diag_kernel).updateMask(directions.eq(2)));     // and add the bands for rotated kernels     for (var i=1; i