topic extrapolation in ArcGIS Pro Questions

extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 11:08:21 GMT

hi every one,I'm trying to extrapolation my spatial data to larger area, but I don't know if its possible and which tool can.

I m working with pollution concentrations ,its follow a specific spatial trend

thanks

Re: extrapolation

XanderBakker — Wed, 25 Oct 2017 11:19:30 GMT

Can you elaborate a little more on what type of data you have, the current extent of the data and how far you would like to extrapolate the data? Also do you have acces to the Geostatistical Analyst extension?

Re: extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 11:45:11 GMT

yes I have access to Geostatistical analysit extension,I have pollutant concentration of an area and I apply the EBK Regression Prediction method from Geostatistical wizard in Arcgis PRO using altitude raster as prediction variables.for the area i got very pretty result but I wanna to extrapolate for larger area.

the current extent of the data

top :50.856194

bottom:50.767833

left:15.151972

right:15.357413 dd

and I like to extrapolate to

top :51

bottom:50.7

left:15.10

right:15.4

please note that I tried the extent tool but it s not activated in arcgis pro

Re: extrapolation

DanPatterson_Retired — Wed, 25 Oct 2017 11:48:19 GMT

Set your analysis extent to the larger extent prior to running the tool. In any event, extrapolation is generally a bad idea unless you are absolutely sure that trends seen at the edges persist beyond the margins

Re: extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 11:50:46 GMT

Yes I m sure that there is a trend it been demonstrate by prior studies in the field.

Re: extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 11:52:27 GMT

yes I m trying to extent this to larger area but I coudnt do it in arcgis pro like I use to do in arcmap

Re: extrapolation

XanderBakker — Wed, 25 Oct 2017 11:56:33 GMT

Just remember that it is a relatively large area that need to be covered by extrapolation. Without some additional information, like a trend or something that would enable Co-Kriging, your result will not be very accurate

Re: extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 12:24:41 GMT

thanks for your answer,but I how can I do the extrapolation which tool I must use

Re: extrapolation

XanderBakker — Wed, 25 Oct 2017 12:37:07 GMT

So, technically you can use extrapolate your data to a larger extent, as long as you follow the instructions provided by Dan (set the desired extent for your output in the Geoprocessing Environment). Which tool can you use? Probably any interpolation tool available in Spatial Analyst/3D Analyst and Geostatistical Analyst. Since you have Geostatistical Analyst, this is the preferred tool for the job. If you have any other information that has coverage for the entire area and there is a strong relationship between this data and your pollutant concentration, you would use this relation in Creating a prediction map using cokriging—Help | ArcGIS Desktop to create your map. Geostatistical Analyst is also preferred since it will create a probability map that will tell you more about the quality of your result. Expect lower quality the further you are from the area covered by your data.

Also have a look at: What are the different kriging models?—Help | ArcGIS Desktop

Re: extrapolation

AMALBOUZIANE — Wed, 25 Oct 2017 12:43:51 GMT

great thanks

Re: extrapolation

SteveLynch — Wed, 25 Oct 2017 15:28:08 GMT

You cannot set the extent larger than that of the input points with EBKRegressionPrediction. The other interpolation tools (except NaturalNeighbor) can be used for extrapolation.

However, be warned. It is a bad idea to extrapolate.

Re: extrapolation

DaniloDurante — Tue, 02 Apr 2024 17:59:41 GMT

Here are a few solutions listed:

https://support.esri.com/en-us/bug/unable-to-extrapolate-the-values-on-a-geostatistical-an-bug-000109393