8

I am using ArcGIS Desktop 10.1.

I am trying to find the location of the highest value pixel in my Direct Solar Radiation Raster. I can see the extent of the data in my raster but don't know where the highest value pertains to.

Is there a way to simply select pixel with highest value or create a point at highest value?

PolyGeo
  • 65,136
  • 29
  • 109
  • 338
Calavin
  • 869
  • 2
  • 10
  • 19

2 Answers2

6

From what I know, there is no direct tool. So you need to :

1) read the maximum value in the properties of your raster (right click on the raster > raster properties ) (if not available, launch "build statistics"(management tools > raster > raster properties). Make sure that your skip factor is set to "1" in order to use all pixels during the calculation (skip factor correspond to the pyramid level)

raster properties with Min and Max

2) use the "Equal to" tool (spatial analyst > math > logical) to identify the cells with the maximum value

remark : alternatively you can do this with "reclassify" or map algebra.

3) convert raster to point

radouxju
  • 49,636
  • 2
  • 71
  • 144
  • 2
    (1) Is it not the case that ArcGIS will only subsample the raster data for large rasters when computing the source statistics? If so, it will be unreliable to obtain the maximum in this way. (2) There is a "direct" calculation, although it takes a few steps: compute the zonal max of the grid based on a zone created from the grid itself. – whuber Jan 19 '14 at 23:03
  • Thanks for the remarks, but if your skip factor is equal to one, as I did on my example, you actually use every single pixel http://blogs.esri.com/esri/arcgis/2013/10/11/skipfactor/ . I've edited my post accordingly. 2) zonal statististics might also use some kind of resample internally with very large datasets, to be checked in the environment settings. – radouxju Jan 20 '14 at 06:30
  • (+1) Although I shouldn't be shocked at the prospect of zonal stats using resampling (this software has a long history of strange behaviors that result in incorrect calculations) that would be very, very wrong. Could you provide a more explicit reference that indicates it might do that? – whuber Jan 20 '14 at 16:03
  • With feature dataset, the ArcGIS help says that feature to raster is applied. Then if you look further, you find : "The default cell size is the shortest of the width or height of the extent of the input feature dataset, in the output spatial reference, divided by 250". With raster dataset, it is quite OK but you could have an internal resampling if the resolution do not match. – radouxju Jan 20 '14 at 18:35
2

At Arc Desktop 10.7.1, this works for me (note that it requires the Spatial Analyst extension):

Open the Raster Calculator and enter the following in the Map Algebra Expression pane:

Con("your_raster" == "your_raster".maximum, "your_raster")

[Note: this selects the cell with the maximum value from your_raster.]

Enter the name of the output raster that will contain the cell with the highest value. Voila!, the output raster with the maximum value is created. All other output cells are NoData.

If the output cell is too small to find, use the Raster to Point tool. Then symbolized the output point with a large, brightly colored dot.

Stu Smith
  • 8,252
  • 8
  • 35
  • 84
  • Thanks for this, worked great and is the solution I was originally looking for. Was this always an expression that could be used or is it new with 10.7? – Calavin Feb 12 '20 at 15:44
  • I'm sorry, but I can't definitely answer your question. I'm not sure when I first learned about ".maximum" (although it was probably within the last 12-18 months), so I don't know when it was first released. Also, I've looked through my notes, as well as ESRI help, and a web search, but I can't find anything about it. Weird. I sure wish I had a better memory!!! – Stu Smith Feb 13 '20 at 06:30