Question re: Land Use Classes

[Justin]

Colleagues,
I have many crop rotations in my study area. I designated them
20001, 20002, etc., to differentiate them from Crop Dataset Layer
(CDL) and the National Land Cover Dataset (NLCD) layers. For
instance, a rotation of corn in 2006, wheat in 2007, and corn in 2008
would be designated as class 20001; a rotation of corn in 2006, corn
in 2007, and wheat in 2008 would be designated class 20002, etc.

To ensure 100% overlap between the land use file and the delineated
watershed, I combined the CDL data layers for each year for a larger
mask exceeding the size of the watershed and imported this grid into
arcswat during the "Land Use/Soils/Slope definition." Now, in the
land use lookup table for that mask, I accidentally typed:

20001,PAST
20002, URLD
.
.
.
20065,PASt

where 20001, ...20065 indicate land use class numbers stipulating
different crop rotations (see above). "PAST" refers to the official
pasture land use class, which is the land use class for the initial
year in both locations, but note that while both locations may have
had the same initial land use class, the land use class varied later
on (that's why they are designated with different land use class
numbers).

SWAT loaded this and informed me that:
20001 PAST 69.19%
20065 PASt 0.1%

Correcting my typo by changing the lower-case "t" changes the table
that SWAT produces to say:
20001 PAST 69.29%

Obviously, not every land use class in the lookup table appears in the
chart given during the land use definition since the lookup table was
valid for the larger mask but SWAT only reads in what is valid for the
watershed. Note though, that 20001 and 20065 denote different crop
rotations, even though both areas were pasture in 2006. So, it seems
to me that SWAT is combining different land use classes identified by
different numbers into one larger class. Is there a way around this?
While the land uses for classes 20001 and 20065 may be the same
initially, they diverge later on so I'd like to keep them separate so
I can appropriately apply the land use classes later on. Ideally, I'd
like to know which proportion of my watershed is class 20001, which is
class 20002, etc., assuming that all classes apply in the actual
watershed.

Purposefully mistyping a land use name won't resolve the issue
either. SWAT doesn't know what to do with "PASt," but it does know
what to do with "PAST."

In short, I'm afraid that SWAT may be combining land use classes
based on the same initial land use when I'd rather it not.

Do you have any ideas on how to circumvent this?

Thanks,
Justin

[Jim Almendinger]

Justin --
As you surmise, ArcSWAT uses the initial land-use category in the land-use grid or shapefile in creating HRUs.  It's not using the grid value -- it's using the reclassified SWAT land-use name that you assign to that number.  All the different grid values that you assign to PAST will get called the same thing by ArcSWAT -- at that point it has no idea that management practices will differ in later years. 

The initial land-use designation in SWAT is really not that important for those places where you're going to institute a crop rotation.  Once you start a rotation, SWAT will forget what the initial land-use designation was.  Lately I've taken to lumping all my agricultural land into a single SWAT category (AGRR, agricultural row crops), and then afterward I'll assign different crop rotations to this land (I have more detailed information on crop rotations from county agents that I think are better than using the sequential CDL data layer methods). 

Your options:
(1) Easiest would be to assign different initial land uses (crop types) from the many existing SWAT land uses to your different pasture scenarios.  E.g., call one pasture, call another fescue, call another brome, another bluegrass, etc.  It really doesn't matter what you call it (as long as it's unique and not conflicting with a real land use elsewhere) -- you're going to change it anyway when you enter the rotation information later. 
(2) Probably better would be to add your own crop types to the crop data base -- I think you're allowed to do this.  If you've got 4 different pasture rotations, I'd simply copy the PAST data line 4 times in the data base, and rename each type something sensible you can remember, like PAS1, PAS2, etc.  (I don't know the rules for adding crop type and naming conventions.)  Later you'll have to add the rotation information, but at least this way you'll retain the spatial locations of these different rotations. 
(3) Otherwise -- what I commonly do is to accept the lumping of similar land uses by ArcSWAT, and assign rotations to HRUs later.  If I know I've got 500 ha of Corn-Soybean rotation in the watershed, I'll export the HRU table with areas, sort to find the AGRR HRUs (see note above).  Then I'll use a formula with a random number generator in it to select an array of AGRR HRUs and keep making trial random selections until I get a total area very close to my target.  Then I'll take those Corn-Soybean HRUs and assign half to a Corn-Soybean rotation, and half to a Soybean-Corn rotation. 
     This involves a lot of copying, pasting, sorting, and so forth in Excel, and I've lost some minor control on where these rotations end up on the landscape -- except it does of course crops do end up on cropland. 

Distributing your management scenarios (rotations) among your HRUs is one of the most tedious parts of constructing a SWAT model, I think.  But I'm not sure how to improve the methods. 

Cheers,
-- Jim

---

From: "Justin" <gold...@gmail.com>
To: "SWAT-user" <swat...@googlegroups.com>
Sent: Wednesday, December 1, 2010 8:32:09 PM
Subject: [SWAT-user:2496] Question re: Land Use Classes

--
You received this message because you are subscribed to the Google Groups "SWAT-user" group.
To post to this group, send email to swat...@googlegroups.com.
To unsubscribe from this group, send email to swatuser+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/swatuser?hl=en.



--
BEGIN-ANTISPAM-VOTING-LINKS
------------------------------------------------------

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

Teach CanIt if this mail (ID 01DC2werz) is spam:
Spam:        https://canit.smm.org/canit/b.php?i=01DC2werz&m=f581f675c3fd&t=20101201&c=s
Not spam:    https://canit.smm.org/canit/b.php?i=01DC2werz&m=f581f675c3fd&t=20101201&c=n
Forget vote: https://canit.smm.org/canit/b.php?i=01DC2werz&m=f581f675c3fd&t=20101201&c=f
------------------------------------------------------
END-ANTISPAM-VOTING-LINKS



--
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