thank you very much for your quick answer! 

When downloading or clipping the image, I did not specify anything in the Environment settings, so I assume that the coordinate system is unaltered during the processes.

I had this problem when trying to download monthly climatological raster images of CHL_a (4 km spatial resolution) from NASA Ocean Color (AquaMODIS) using the following tools:

MGET>Data Products>Nasa GSFC OceanColor Group>L3 products>Create Climatological raster ... 

To save you some time downloading the images, maybe I can give you the complete (from the Ocean Color website) and the clipped images (downloaded with MGET). Please, let me know what you prefer and if you have all the information you need to further explore this problem.  

Thank you very much for your help!

Best

Victoria

2017-02-17 11:04 GMT-03:00 Jason Roberts <>:

Dear Victoria,

 

That is an unusual situation. The change in the No Data value is ok and not something to be alarmed about. The increase of the min value from 0.01 to 0.08 is plausible; it may be that the clipped area does not have any values less than 0.08. But the increase of the max value from 35.6 to 95.3 is concerning. Just like the min value should remain the same or get larger after clipping, the max value should remain the same or get smaller.

 

All of that assumes that the coordinate system is not changing somehow during the clipping operation. If the coordinate system is changing—e.g. the data are being projected from the original WGS 1984 “geographic” coordinate system to something else—then it is possible for this situation to occur.

 

I have not seen this happen when the coordinate system has not changed. Can you please give precise instructions on how I can recreate this problem? If so, I can look into it further.

 

Best,

Jason

 

From: Victoria Gonzalez Carman [mailto:]
Sent: Thursday, February 16, 2017 10:29 AM
To: Jason Roberts <>;
Subject: clipping a raster downloaded with MGET changes no data and maxmium values of aqua_CHL_chlor_a

 

Dear Jason, how are you? I hope you are doing well. 

 

I used MGET to download raster images of surface chlorophyll clipped at the Southern Hemisphere. I realized that the clipped images have ranges of values that are different from the original complete images (in NetCDF format) downloaded from the Ocean Color website. Besides, the no_data values also change. 

 

For example, 

 

Original raster image (not clipped):

range of values (min - max): 0.01 - 35.6 mg/m3. 

no data: -32767

 

Clipped images

range of values (min - max): 0.08 - 95.3 mg/m3. 

no data: -3,40282346639e+038

 

What is weird is that the maximum values increases a lot, and that the nodata value also changes. 

 

I have read in some forums that maybe there is something wrong with the resampling procedure used by arcmap, so I tried to change the re-sampling options in the Environment section but I continue having the same problem. 

 

I also tried re-scaling the raster, but now the original and the clipped image does not look like the same using the same symbology options, which it makes me feel insecure about this procedure. 

 

I will appreciate your help with this problem, and I apologize if my question is too simple.  

Best regards

 

--

************************************************************
Dra. Victoria González Carman
Instituto de Investigaciones Marinas y Costeras (IIMyC - CONICET)

Proyecto Ecología pesquera - INIDEP
Aquamarina- PRICTMA
Mar del Plata, Argentina. 

http://www.iimyc.gob.ar

http://www.ecologiapesquera.com.ar

Skype: vgcarman
************************************************************