Help - Selecting minimum value out of multiple raster layers

[ToniRamirez]

I am calculating the Newmark displacement (ground failure) for my local area. I calculate this in various percentiles (1, 10, 30, 50, 70, 90, and 99) creating multiple raster layers that overlap. Some cells will have multiple values because of the different percentiles. I am needing  the minimum value within each cell out of the raster layers and to create a singer raster file if possible. Can someone help me with a tool or way to achieve this outcome? 

Thanks in advance!

[DanPatterson · Accepted solution]

General Spatial Analyst toolset help

An overview of the Spatial Analyst toolbox—ArcGIS Pro | Documentation

You need  a "Local Minimum" which falls under Cell Statistics

Cell Statistics (Spatial Analyst)—ArcGIS Pro | Documentation

[ToniRamirez]

Thank you that worked! 🙂  Is there a way of knowing which raster dataset was selected for the minimum used. I am having trouble with the classification for the symbology.  

[DanPatterson · Accepted solution]

Of course ... see An overview of the Local toolset—ArcGIS Pro | Documentation

for example

Lowest Position (Spatial Analyst)—ArcGIS Pro | Documentation

Please "Accept as Solution" the response that answered your question so that the thread is closed with a known solution.

[ToniRamirez]

Hi @DanPatterson , 

I was hoping you could help me on this problem again. Using the Lowest Position tool, if one of the rasters has a cell with "No Data" it returns the output cell with no data. Is there a way for it to choose the lowest value available from the input rasters rather than the cell of the no data value?

[DanPatterson]

not unless you convert all your rasters to numpy arrays... which has "nan" functions to account for such situations.

reclass is possible... reclass all you nodata values in all arrays to the highest value allowed by the raster data type  (eg 2**8, 2**16 etc

If a cell location contains NoData on any of the input rasters, that location will be assigned NoData on the output.

[ToniRamirez]

Okay, I will gives these a try. 

Thank you!