aggregate daily values to monthly for each pixel in a polygon using google earth engine

Question

I am processing the NASA NEX-GDDP data on google earth engine for a polygon. This data has three varibales: pr, tasmin, tasmax and is daily resolution. For each day and variable, there are 21 images till 2005 and 42 from 2005 onwards.

My goal is to calculate monthly sum (for pr) and monthly mean (for tasmin and tasmax) in each year for each variable for each grid located in the polygon for every single image in a day. Therefore, for a given variable, a single row in my final output should look like:

 lat    long     month    year   ACCESS1-0' 'bcc-csm1-1'  'BNU-ESM'  'GFDL-ESM2M' ...... 'NorESM1-M' 
 8.125  108.875   1       1950   350          380           320         333                 322

The approach I am adopting is to write out daily values for each lat lon for each variable and do the summing up in R. In order to write out daily values, I used the following approach

 // define which variable I want 
 var myvar = 'pr'

 // define my polygon
 var polygon = ee.Geometry.Polygon([120, -6, 120, 8, 99, 8, 99, -6]);

 // Define a start and end date
  
 var startDate = ee.Date('2006-01-01');
 var endDate = ee.Date('2018-01-01');
  
 // Filter the image collection by date range, polygon and variable
 var dataset = ee.ImageCollection('NASA/NEX-GDDP')
               .filter(ee.Filter.date(startDate,endDate))
               .filter(ee.Filter.bounds(polygon))
               .select(myvar);
  
 var dailyImg = dataset.toBands();

 var collection = dailyImg.sample({
   region: polygon,
   geometries: true,  
 });
  
 // Break point coordinates up into properties (table columns) explicitly.
 var collection_with_latlon = collection.map(function (feature) {
 var coordinates = feature.geometry().transform('epsg:4326').coordinates();
   return feature.set('lon', coordinates.get(0), 'lat', coordinates.get(1));
  });
  
Export.table.toDrive({
   collection: collection_with_latlon,
   description: 'tx_2006_2017',
   fileFormat: 'CSV',
 });

Each row in the resulting file is a lat lon from the polygon and the days are arranged as columns. So the number of columns I have is number of days * 21 GCMs. This table can get very big column-wise if I increase the number of days

How can I write my output such that each row is lat/long and a day combination?

aggregate daily values to monthly for each pixel in a polygon using google earth engine

0 Answers0