Calculating Area on Google Earth Engine

[@prof.biplabbiswas5576]

EXCELLENT!!

[@bastianalejandrorivasmaldo1217]

Hola Solve Scholar, podrías subir un video de como calcular area para series de tiempo modis?

[@yogofa]

Thanks for your informative video. You claim that pixel areas are calculated according to Lambert Azimuthal equal area projection. However, if you type print(ee.Image.pixelArea().projection()) in GEE code editor, you will get a WGS84 projection (EPSG:4326). That being said, I don't know what is the correct projection of ee.Image.pixelArea()! Best Regards

[@SolvedScholar]

One thing is the projection in which the m² values are calculated. Another is the projection in which a given layer is represented on the canvas. Regarding the fact from where this info comes from, it comes from inside the Google Developer team itself, according to Mike Dixon (one of Earth Engine Developers). On 29 January 2019, I've asked the following question: "...from the cartographic point of view, what does ee.Image.PixelArea () do?" Mike Dixon's answered: "The current algorithm is: 1) Construct a Lambert Azimuthal Equal Area projection centered near the pixels (when we're processing a block of pixels, we try to use a single projection centered on the block, but if the block covers too large an area we subdivide it and use separate projections for each piece). 2) Transform the four vertices of each pixel from the current projection into the LAEA projection. 3) Compute the area of the resulting quadrilateral using the standard formula for the area of a polygon from its coordinates". Want to interact with specialists in Google Earth Engine...subscribe to the Google Earth Engine Developers Group.

[@yogofa]

@@SolvedScholar Thanks for your time and effort in answering my question. If I reproject the ee.Image.pixelArea() image to another projection, will the pixel areas be calculated according to the newly specified projection or to LAEA?

[@SolvedScholar]

@@yogofa You will never be capable to change the way the ee.Image.pixelArea() was develop to work. The ee.Image.pixelArea() is a function not an image. This function has no parameter allowing you to modify the calculation projection. Thus there is nothing you can do about it.

[@yogofa]

@@SolvedScholar I'm grateful for you. You have clarified a lot of ambiguity for me.

[@INSGIS]

Hi Ceaser, I have a problem with this topic. I exported my image to my drive, then I converted that image using ‘Raster to Polygon’ tools in ArcGIS without 'Simply Polygons'. I have different calculation results between EE and ArcGIS. Even though I have projected the image and its calculation using 'EPSG:32749'. How did this happen? can you give me an explanation?

[@SolvedScholar]

Dear @INS GIS, this can be related to several causes, without seeing your full code there isn't much I can do. But this topic here may help you out, have a look: groups.google.com/forum/?#!msg/google-earth-engine-developers/l1SSbicyEFk/uHpv0F2ZBQAJ Regards Cesar

[@INSGIS]

@@SolvedScholar Thanks Ceaser 🙏

[@NgocNguyen-ys8ec]

uhm, can you compare between Reducer and gt/eq/lt? As I know, all of them are used to mask out the unwanted region

[@mamponsah6213]

Thanks for the lecture. Please, what does 1e13 in the line " .reduceRegion(ee.Reducer.sum(),geometry,10,null,null,false,1e13)" stands for Thanks

[@SolvedScholar]

It is an argument of the reduceRegion function, it referes to maxPixels (Long, default: 10000000) and sets the maximum number of pixels to reduce.

[@mamponsah6213]

@@SolvedScholar Thank you

[@janethchaparro8017]

Excellent contribution. How can I calculate the area by class in an unsupervised classification?

[@SolvedScholar]

The ideia isn't different...you will need to use the ee.pixelarea as usual

[@Luke_o_lophus]

could you suggest how to calculate the areas of each feature in a supervised landcover classification? Thanks in advance! The roi was obtained from a featurecollection and several land use types under property 'landcover'

[@SolvedScholar]

The ideia isn't different, the classifier type does not metter...you will need to binarize your classes and use the ee.pixelarea as usual

[@dhritirajsengupta7888]

Great Video! thanks a lot. Just one question. I am working with time series data and using ee.Image.pixelArea, is there any overlapping of area when I move from 1990 to 1991? I am calculating annual change in vegetation .

[@SolvedScholar]

You have to temporally filter your data. Let's say you are working with median composites, temporal overlapping will only happen if you set it to happen, for instance, in the .filterDate

[@dhritirajsengupta7888]

@@SolvedScholar code.earthengine.google.com/f3048af802cd22dc1b4ff494430609ca this is the code i am working with. Could you please look at it once?

[@SolvedScholar]

@@dhritirajsengupta7888 you forgot to share your assets

[@syedmohammadmuzammil7118]

@@dhritirajsengupta7888 Hey! did you get the answer? I am also working on something like this. could you help?

[@SolvedScholar]

@@syedmohammadmuzammil7118 Hi Syed, it will be a pleasure to help you out. Please share your code and assets if necessary. All the best.

[@omerreshi]

Respected admin, I have generated Ndvi , want to calculate the area of vegetated portion and non vegetated portion. How to do this.

[@SolvedScholar]

First you need to classify the vegetated area, binarize your classification and then calculate the area.

[@syedmohammadmuzammil7118]

Can you help me to calculate the change in the area in time series chart?

[@SolvedScholar]

The ideia isn't different...first you will need to detect the difference. Then, you will need to use the ee.pixelarea as usual

[@camilovillamercado684]

Can you help me, How can I calculate the area of ​​urban areas, some type of sum of areas of some combination of bands?

[@SolvedScholar]

First, you need to classify your "Urban Area". Have a look at the topics of "impervious surface mapping" and "night light detection sensors". Once you classified it, send us the code.

[@edwinosei1332]

Is it possible to convert the km2 to percentage? And how is it done?

[@SolvedScholar]

Yes it is. But percentage in terms of what? Total area of the country? State? Municipality?

[@edwinosei1332]

@@SolvedScholar of the pixelarea

[@SolvedScholar]

@@edwinosei1332 the ee.ImagepixelArea is a pixel based operation. It will store the pixel area as a single m2 value. There is no half, thirds or quarters of pixels being considered. That said, to convert the m2 value to percentage, you will need to dived the area of your class by the total area of an existing reference. For instance, the percentage of agriculture surface in the state of Oregon, will be to total of agriculture labeled pixels divided the total area of the Oregon state itself.

[@edwinosei1332]

@@SolvedScholar thanks

[@jhonmoreno4593]

Thanks for de video! Id like to calculate areas from NDVI but I dont know how to binarize my image. Please, Can you help me with any idea to achieve it?

[@nishantthaploo2539]

Hi, I'm getting an error which says , Dictionary.get: Dictionary does not contain Key: constant.......its some sort of number(error), i hope i could get some help...

[@josebofana6315]

Thanks for the video, really useful. However, I have followed it to calculate the area of the cropland extent in Brazil, but there is an error coming out on the GEE as following: Number (Error) Computation timed out. Any suggestion on how to solve it? thanks in advance.

[@SolvedScholar]

Computation time out can be caused by many reasons, can you send me your script?

[@josebofana6315]

@@SolvedScholar code.earthengine.google.com/c8d7cd581e64ed33d8290226f3dda7ea?fbclid=IwAR2gTA9bqcO_SOyf44ShXkjhmule_BzxNqmcsmE47wDCC_BDj6Vy7S_6c74 Here you have...

[@SolvedScholar]

@@josebofana6315 here it goes: code.earthengine.google.com/0e485c465fc29c87a03096ec41734995?fbclid=IwAR2gTA9bqcO_SOyf44ShXkjhmule_BzxNqmcsmE47wDCC_BDj6Vy7S_6c74

[@osm4392]

Where are you from? It seems you are trying to emulate a british accent. Trying too hard that it sounds too irritating and too difficult to understand some of the words!!!

[@osmarcuentastoledo384]

It´s interesting