Editing SSURGO/STATSGO database to write .sol table

[Adam Backsmeier]

I am seeking assistance to overcome discrepancies between my SSURGO/STATSGO data downloaded from the Geospatial Data Gateway and the SSURGO/STATSGO databases available for use by ArcSWAT.  This issue arose when attempting to write the .sol table.

I am using ArcMap 10.3 and ArcSWAT 2012.  I have already downloaded the SSURGO database from the ArcSWAT website and moved it to C:\SWAT\ArcSWAT\Databases folder. When using STATSGO soil data, I utilize the database included with the ArcSWAT download.

I am aware based on this post that I should append the database using Microsoft Access, but I am not sure where to obtain detailed information regarding specific soil Map Unit Keys (mukeys).  Likewise, I am not sure where to find specific soil data for STATSGO soils.

Without resources available to input data, I have attempted to remedy the problem by searching for soils with similar mukeys and replicating the data in the database for the missing mukey.  For example, the SSURGO database from the ArcSWAT website is missing mukey 2592231.  I created a new entry in the SSURGO database for this soil using the data from mukey 2592230.  The assumption is that the soils are sufficiently similar to negate any impact on the model.  I used the same methodolgy for STATSGO.

I am not pleased with this solution, especially since there are several discrepancies, so I am hoping a better method exists.

Requests for help:

1) Source of information for specific mukeys that can be used to fill-in missing data in soil databases.

2) Alternative method of overcoming discrepancies between SSURGO/STATSGO data and databases.

This is a great community and benefit to SWAT users.  Thanks for the help.  

-Adam

[Jim Almendinger]

If I recall, the STATSGO data set downloaded with ArcSWAT already has selected a soil series, with all the relevant data needed by SWAT, to represent each soil mapping unit (a group of soil series, more or less) given in the STATSGO grid.  Mostly you need to make sure that the SSURGO database downloaded with ArcSWAT includes the soil types (identified with the mukey or muid) in the STATSGO part of your data set.  In your STATSGO grid, you may need to change the name of the muid field to mukey (or vice versa -- I forget) -- you want it to be consistent with the SSURGO terminology.  You will want to mosaic all of your soil layers (probably all grids now?) together, so that there is a single soils grid with every cell having an muid value that is in the SSURGO database.  This is your goal, prior to trying to build the model. 

The soils databases are in the installation folder, probably on your C: drive.  If I recall, the SSURGO database is nested in its own folder, and you may need to drag it up one directory level into the main "swat databases" folder.  That way ArcSWAT can find it. 

If the SSURGO database does not have the specific soil type in your STATSGO data set, you need to find a substitute.  By far the most critical decision here is to make sure you replace it with a soil of the same hydrologic soil group (HSG), which is a primary determinant of curve number, which controls the infiltration vs. runoff calculations in SWAT. 

This is from memory, so some details may be a bit wrong...

Good luck,

-- Jim

---

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arcswat+u...@googlegroups.com.
To post to this group, send email to arc...@googlegroups.com.
Visit this group at https://groups.google.com/group/arcswat.
For more options, visit https://groups.google.com/d/optout.

---

Spam
Not spam
Forget previous vote

--

Dr. James E. Almendinger
St. Croix Watershed Research Station
Science Museum of Minnesota
16910 152nd St N
Marine on St. Croix, MN  55047
tel: 651-433-5953 ext 19

[Adam Backsmeier]

Jim,

Thanks for the detailed reply.  I was able to complete all the steps you outlined, but I am not sure where to find information on acceptable substitutes.  I have approximately 95 SSURGO mukeys that do not exist in the SSURGO database from the ArcSWAT website.  I would simply add them to the database, but I am not sure how to make acceptable substitutions. What are your thoughts?  Thanks.

-Adam

[Jim Almendinger]

I don't remember having so many missing soil types; 95 seems like a lot.  Are these from the STATSGO data set, and if so, aren't the soil properties for those types already in the STATSGO table, that you could just copy to the SSURGO table? 

If the unknown types are from a newer SSURGO layer that you used (and therefore unrelated to the STATSGO issue), with types not in the SSURGO table downloaded with ArcSWAT, then you need to pick similar soils.  The most critical point is to get the hydrologic soil group the same -- after that, you can choose other soils data (depth, soil K, etc.) as secondary criteria, if you wish.  Otherwise I think I'd just find the most prevalent soils in each of the HSGs (A, B, C, and D) and assign those to the unknown soils. 

-- Jim

---

From: "Adam Backsmeier" <aback...@gmail.com>
To: "ArcSWAT" <arc...@googlegroups.com>

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arcswat+u...@googlegroups.com.
To post to this group, send email to arc...@googlegroups.com.
Visit this group at https://groups.google.com/group/arcswat.
For more options, visit https://groups.google.com/d/optout.

---

Spam
Not spam
Forget previous vote

[Adam Backsmeier]

Jim,

Thanks for the reply.  I have pulled up the STATSGO and SSURGO soil databases to try to fill in the missing mukeys, but I am not sure how to make the connection between the databases.  The SSURGO muid is a 5-digit number while the STATSGO muid is a 2-digit state abbreviation and a 3-digit number (e.g. WY010). Additionally, none of the S5IDs or muid+seqn are able to be matched.  What method would you recommend for matching up soils in both databases to fill in the missing mukeys?  Thanks.

-Adam

[Jim Almendinger]

As I look back through my notes, I think I didn't use the ArcSWAT downloaded STATSGO data -- I downloaded STATSGO2 from the web, and I think it makes a more direct connection with the SSURGO naming conventions.  Every soil type in my STATSGO2 data set was already in the SSURGO database. 

I agree, that when I look at the SWAT-downloaded SSURGO and STATSGO datasets, I don't see any obvious key to relate them together. 

So that's my advice at the moment -- use STATSGO2 rather than the ArcSWAT-downloaded STATSGO dataset.

-- Jim

---

From: "Adam Backsmeier" <aback...@gmail.com>
To: "ArcSWAT" <arc...@googlegroups.com>
Cc: "Adam Backsmeier" <aback...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arcswat+u...@googlegroups.com.
To post to this group, send email to arc...@googlegroups.com.
Visit this group at https://groups.google.com/group/arcswat.
For more options, visit https://groups.google.com/d/optout.

---

Spam
Not spam
Forget previous vote

[Adam Backsmeier]

Jim,

I really appreciate your assistance as I work through this issue.  I went back to review my data and confirmed that I was using STATSGO2 data downloaded by state from the Web Soil Survey.  My watershed straddles a state boundary so I merged and dissolved the soil data based on mukeys.  I also projected in the appropriate projection for my model data.  

When using gSSURGO, 89 mukeys are missing as explained above; however, when I try with STATSGO2 data, a different problem arises.

When I am writing the .sol table, a dialog box appears saying that soil WY267 (or any other muid, such as ID101) does not exist.  When I look up the STATSGO database organic to ArcSWAT, the stumid is present with several sequence numbers (e.g. muid: WY267 seqn: 2).  The problem is that when I load my soil shapefile from the disk and select STATSGO>stumid+seqn, the sequence numbers do not appear. 

As such, I am left with missing SSURGO mukeys and a communication error between ArcSWAT and the STATSGO database.  Any recommendations to get past my soil issues?

Thanks again for the dedicated help!

-Adam

[Regensburger, Eric]

Adam, I had a similar problem with the ArcSWAT STATSGO database regarding a project with soils from more than one state. Even though the soils I was using were in the ArcSWAT STATSGO database, ArcSWAT would only recognize the soils from one state (Montana) in my model boundary. I got around this problem by appending the Montana specific ArcSWAT database with the soils that I needed from the other states (ID and WY). I kept the soil names and all information the same from the WY and ID databases and just added them to the end of the MT database – that trick worked for me.

--

[Jim Almendinger]

Personally I'd try to make the SSURGO soils data work, because the spatial resolution is so much better.  I've been burned a bit with STATSGO soils not being representative at the scale I'd hoped for.  Your task will be to find appropriate replacements for the 89 missing soils.  Replacing them with soils of the same hydrologic soil group (HSG) is the key here. 

Eric's suggestion for STATSGO soils is a good one.  I don't know why ArcSWAT didn't find the soils from the other states -- perhaps it's just limited to reading soils data from a single table.  But each time I've had to piece together soils data, I created (or modified) a single database table with all the types in my spatial dataset, which has worked every time.  When I did this the last time, one county had only STATSGO data -- so I merged the STATSGO2 polygon shapefile with the SSURGO polygon shapefile.  And the MUKEYs in both sets corresponded to the MUIDs in the SSURGO table supplied with ArcSWAT, so I had no missing values. 

Beyond this I'm not sure what to recommend. 

Best of luck,

-- Jim

---

---

NOTE: This message was trained as non-spam. If this is wrong, please correct the training as soon as possible.

Spam
Not spam
Forget previous vote

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arcswat+u...@googlegroups.com.
To post to this group, send email to arc...@googlegroups.com.
Visit this group at https://groups.google.com/group/arcswat.
For more options, visit https://groups.google.com/d/optout.

[Adam Backsmeier]

Eric,

Thanks for the recommendation.  When you say append, do you mean something more than copying and pasting the information from one state database into another.  That is what I did and it did not work.  Unfortunately, I am not familiar with Access so your help is greatly appreciated.  Thank you.