I have a dataset that ranges from 2016 to 2021. The file types that I am going to use for the analysis are Aqua Modis and NetCDF-4. My data contains 650 rows and I am using parameters such as sea surface temperature, sea salinity, distance from shore, and others to configure which biophysical aspects affect the distribution of dolphin species.


My aim is to conduct a multiple overlay analysis of data layers containing the values for sea surface temperature per day, per month, per year, superimpose and merge all these layers together to produce a single map layer in order to calculate the average sea surface temperature for all our research days per row in my .csv datasheet from the individual multiple data layers ranging from 2016 to 2021. I need to extract these values for statistical analysis in R software.

I am a complete novice to QGIS and I cannot figure out how to solve this conundrum.

1 Answer 1


I think you've posted more or less the same question twice. (No need to repeat: On StackExchange answers are usually very quick when the question is well focused).

In general, the challenges that you bring up can be accomplished both in QGIS, and in R, and both software can, in general, read NetCDF files and extract raster values (i.e. SST) at point locations. And both can prepare time series.

I think that the main challenge you might not be aware of is the un-projected nature of MODIS data. The SST datasets from, for example, the NASA Oceancolor site, are matrices of values with no geo-referencing, and along side, two equivalent matrices of latitude and longitude for each of the values. User's need to read the values, and the long/lat and then create a coordinate reference system and project (and interpolate) the values onto that CRS.

This can be done manually (coded in python or R, for example) or there are a couple of tools out that make this easy:

  1. First, NASA now the "Giovanni" platform where you can choose, display and then save to Geotiff a wide variety of Earth Observation data from NASA, including SST. This quickly gives you a global dataset at one time slot, and at 10 km resolution.
  2. Second there is "SeaDAS" an addon to the ESA (European Space Agency) tool known as SNAP. This tool is relatively easy to use, and can reproject and save to GeoTiff. Furthermore you can prepare scripts to loop thru all your hundreds of MODIS AQUA images, and process all of them.

So, your project sounds quite interesting and challenging. If I'm not mistaken, it seems you should invest the time to learn all the tools you'll need. Then, once you have choosen which software, and have made some progress, post back if you run into specific problems.

  • Thank you, Micha. I am really a complete beginner with this type of analysis and I have been sitting at my desk for two days trying to figure this out. I initially wanted to import all of my Aqua Modis files into QGIS, convert them to rasta files, overlay them, and then calculate the average SST per ID per Modis file per row over the 5 years period. I deleted the other question and rewrote this one as it is more focused. Commented Apr 4, 2022 at 6:58
  • I have felt confused and I needed the advice to point me in the right direction. One question is focused on doing this analysis in QGIS and the other is doing this analysis in R. Thanks for the advice, you have helped. Commented Apr 4, 2022 at 6:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.