Weighted analysis with multidimensional data

[BettinaStoehr]

In ArcMap 10.4 (advanced licence) I am trying to weigh temperature data in a netcdf-file according to landuse in a tif-rasterfile, and then aggregate that data to the areas of EU-NUTS3-regions, and I am searching for the right spatial analyst tools to do that. I haven't managed to read the NetCDF-file to ArcGIS correctly as a rasterlayer, but I didn't manage yet to add the layer to a mosaic layer, so until then even the spatial analyst tools that will work on a simple rasterlayer will be a great step forward.

I have a

• netcdf-file with EU-temperature-over-time data (time being every day of several years), or single time slices of that file as rasterlayers, if computing NetCDF is too difficult

• a tif-rasterlayer with EU landuse data (industry, inhabited areas, agriculture and other vegetation/landcover)

• a shape with EU NUTS2-regions.

I need to extract a table with the temperature for each NUTS2 region for each day of the year 2010, with the condition, that the temperature has to be weighted according to thermal loads. That means, depending on how much of the area of the temperature-rastercell is covered with industry or other inhabitation, the temperature gets more or less weight (for example, industry can get a weight of 2, normal inhabitation a weight of 1, everything else 0).

My plan was to extract the year 2014 out of the whole range, put it in a mosaic dataset, then with some geospatial analyst tool weigh the landuse data on each of the temperature raster (the netcdf raster is larger than the landuse raster) and then make a statistical analysis of that on the NUTS3 areas. If the Mosaic Part does not work, then extract the 365 timeslices for 2010 as rasters and compute them singularly.

I tried local, environment and overlay on simple rasterlayers, but they did not work. Can anybody recommend the right tools for the analysis?

Thanks and Best regards

Bettina

[BettinaStoehr]

Dear Community,

sorry that I can't ask anyone else - we don't have support at the moment, renewing it slipped through the timepressuregaps. If it helps, I figured that at some stage I have to convert raster to point layer and clip that with my NUTS2 regions. But the weighing problem is still weighing hard on me...

Thanks for anyone taking it's time to help. Please give me a hint if my question is not stated clear enough or sounds chaotic.

Best regards

Bettina

[DanPatterson_Retired]

Bettina, you have many related threads, but you don't indicate if you have examined the tools with arcmap that handle netcdf data and the multidimensional tools that are there.  look at one of you other posts

https://community.esri.com/thread/200044-how-to-extract-a-time-interval-from-netcdf-data-in-arcmap-d...

[BettinaStoehr]

Hi Dan, yes, my posts are all related. I broke the big question down into several smaller ones in the hope to get help with them.

I've been researching thoroughly for over a week (at least as well as I could), I tried the toolbox, I downloaded the multidimensional extension, I found the solution to a silly bug, where arcgis reads my windows7-time-settings, I can read the netcdf-file now as raster and point(=vector)-file correctly, I watched several youtube-videos and the esri-video from 2013ish - I can't watch the 2hour tutorial, unfortunately, because of our licence somehow.

But the tools don't seem to be for analysis with other rasterlayers, or for extracting files.

There is one tool for extracting a timeslice as tif-rasterfile, but it relies on bands that I can't manage to put up.

I also thought that I had rephrased the above question to simply ask for the spatial analyst tool, where I can find out, which % of each landuse is living below my rastercells... Too tired now, but will try tomorrow afresh.