Flow Paths Not Corrected by Cut Lines

23 views
Skip to first unread message

Kurtis Fisher

unread,
Mar 30, 2023, 10:10:36 AM3/30/23
to ACPF Forum
Hello,

I'm brand new to the community and giving ACPF a shot for correcting flow paths in my watersheds of interest. I've been using the Youtube instructional videos thus far and have just completed the cutlines creation step. When I used the manual cutter and dam builder tool, I got my updated flow direction and accumulation rasters and thought these would effectively detect the changes I made with my cut lines. However, the next step of defining an 'area threshold flow network' did not seem to follow my cut lines.

The attached picture (Dap_BS) shows the new flow path still as it was in the original, 'visualize flowpaths' process. However, you can see under my cutlines (I made them transparent) that the DEM was essentially lowered so that flows should have followed their routes. In the additional picture (SumD_LFF) it doens't seem obvious if the elevations were "lowered" but once again, it just didn't work and the water is crossing the road and not the culvert.

Is there anything obvious that I'm not accounting for or a common misstep that anyone might know of in this process?

Thank you,

-Kurtis Fisher
FlowPathNotCorrected_SumD_LFF_DEM.jpg
FlowPathNotCorrected_Dap_BS_DEM.jpg

Asell, Andy

unread,
Mar 30, 2023, 11:32:16 AM3/30/23
to Kurtis Fisher, ACPF Forum
Hmm, you didn't have a subset of your cutlines selected when you ran the Manual Cutter tool, did you? Another thing that looks kind of strange is your hillshade. It looks awfully "stair-steppy" (I'm having visions of rice paddies in China). It appears whatever algorithm used to generate the DEM/Hillshade didn't smooth out the elevations. Thus the reason you have a lot of "straight-shot" lines in your drainage network because they are running over areas of constant elevation. This would affect the viability of the TauDEM extensions when they're trying to model sheet flow. So the underlying original DEM might contribute to the wonkiness of some of your flowlines. But it doesn't answer your original question of why your cutter tool didn't take. Would you be able to reopen your last run of the Cutter too from the results tab and take a screenshot showing your inputs/outputs? Also, What does your updated "Identify Impeded Flow" based off your NewDEM look like after you've run the Manual Cutter? 

ANDY ASELL Environmental GIS Specialist

Water Quality Improvement Section

Iowa Department of Natural Resources

p 515-250-3303

502 E 9th St, Des Moines, IA 50319



--
The ACPF Forum is an online group for discussing technical issues related to using the ACPF tools. Anyone can view the discussions at https://groups.google.com/a/umn.edu/d/forum/acpf-group. Only members can participate in discussions. Join the group by sending an email to acpf-group...@umn.edu
---
You received this message because you are subscribed to the Google Groups "ACPF Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to acpf-group+...@umn.edu.
To view this discussion on the web visit https://groups.google.com/a/umn.edu/d/msgid/acpf-group/19d4e3e0-7bd3-4bdc-ac2e-f952d17e2084n%40umn.edu.

Kurtis Fisher

unread,
Mar 30, 2023, 1:32:29 PM3/30/23
to ACPF Forum, Andy Asell, ACPF Forum, Kurtis Fisher
Hi Andy,

No, I didn't have a subset of lines selected. I actually made sure to clear all selected features at the end of each process for the sake of taking screen shots. 

I'm not sure why that would be the case for my terraced-looking hillshade... I'm using a 5 meter resolution DEM which I generated from a 1 meter DEM with a pretty straightforward cubic resampling method in ArcGIS. I just redid the DEM Preparation steps with an unaltered 10 meter DEM. It ended up looking the same way, just with less resolution. That's an interesting note about my straight-shot lines though, I didn't know it wasn't supposed to work that way... Both DEMS came straight from The National Map, 3DEP quality.

As far as the impeded flow, my cut lines did seem to make them disappear. So, my cut lines worked in a way, but I also need the flowlines to reflect these changes as I'll be using them to correct watershed boundaries in the catchment delineation process. Attached is the picture of my processing results.

Thanks,

-Kurtis Fisher
FlowPathNotCorrected_ManualCutterDamBuilderResults.jpg

David James

unread,
Mar 30, 2023, 1:44:55 PM3/30/23
to Kurtis Fisher, ACPF Forum, Andy Asell
Hi Kurt
A couple of questions...
- Why degrade the DEM to 5m from 1m? 
- The ACPF uses a HUC12-based data structure that uses specific naming constructs. It is best used in this way
- The stair-step appearance that Andy noted is often from raster projection that may not be done well
- Make sure you are using the NewDEM as the argument to the Flowpath tool

DAve JAmes

amarch...@gmail.com

unread,
Mar 30, 2023, 1:49:00 PM3/30/23
to Kurtis Fisher, ACPF Forum, Andy Asell
I’m not sure but maybe the name of your files might be an issue here. Also D drive is not recommended. You should try to run it on the C drive. I’m not sure what is the issue with your DEM but the hillshade looks off to me too. 

Sent from my iPhone

On Mar 30, 2023, at 1:32 PM, Kurtis Fisher <kjfi...@catamount.wcu.edu> wrote:

Hi Andy,
To view this discussion on the web visit https://groups.google.com/a/umn.edu/d/msgid/acpf-group/5592adcd-4845-4cc8-acf3-280e4ac14c15n%40umn.edu.
<FlowPathNotCorrected_ManualCutterDamBuilderResults.jpg>

Asell, Andy

unread,
Mar 30, 2023, 2:04:38 PM3/30/23
to Kurtis Fisher, ACPF Forum, amarch...@gmail.com
You definitely want to go back to the 1 m DM. If you’re worried about processing time, degrading the resolution of the DEM is not the way to go. I’m not sure what your Z units are, but if you can get them into integerized centimeters, your computer should process just fine. If your original DEM add are in float meters, convert to cm by INT( [Input DEM]*100). Once you’re using the new DEM, anywhere in ACPF where it asks for a Z factor conversion, you would set that is 0.01.
--

Kurtis Fisher

unread,
Mar 30, 2023, 2:28:55 PM3/30/23
to ACPF Forum, amarch...@gmail.com, ACPF Forum, Andy Asell, Kurtis Fisher
amarch...@gmail.com,

D drive is my internal solid state drive. Is this not what is meant by a local drive? Should this always be C Drive, no matter what?

Kurtis Fisher

unread,
Mar 30, 2023, 2:35:36 PM3/30/23
to ACPF Forum, Kurtis Fisher, amarch...@gmail.com, ACPF Forum, Andy Asell
My impression was that 1 meter or above is too detailed for watershed delineation, and I thought it would made sense to work within the same resolution as the one in which I will delineate my watersheds to. Degrading to 5 meters from 1 meter resolution was actually a method I used to enhance my 10 meter resolution. I have been warned that 1 meter resolution will be too fine-scaled and will cause issues in the watershed delineation process. Correct me if I'm wrong.

Kurtis Fisher

unread,
Mar 30, 2023, 2:40:33 PM3/30/23
to ACPF Forum, Kurtis Fisher, amarch...@gmail.com, ACPF Forum, Andy Asell
I see now that both amarch...@gmail.com and Dave James pointed out my naming convention. Will this actually cause processing issues? I'm not working from watersheds within the ACPF database so I don't have a Huc12 folder to work from like in the videos.

Asell, Andy

unread,
Mar 30, 2023, 3:06:32 PM3/30/23
to Kurtis Fisher, ACPF Forum, amarch...@gmail.com
1 m is ideal. But if your elevation units are in decimal feet or decimal meters, processing will suffer. Keeping your units in decimal form and increasing the resolution will help processing time, but it will also degrade your results. Converting your 1m DEM with decimal feet or meters Z units to cm is the tried and true way to go. 
--

Agnieszka Marchlewska

unread,
Mar 30, 2023, 3:07:02 PM3/30/23
to Kurtis Fisher, ACPF Forum, Andy Asell
Kurtis,
The C drive is where your Windows is installed and where you store your ACPF too and data. I'm not sure how your computer is set up but D drive is usually set as the network drive and storing the data on the network drive is not recommended becouse it can cause processing errors. The naming convention is very important. stick to what you see on the training videos. Also I noticed that your directory path is very long. Keep it simple. Just create on C drive your ACPD data folder and put your geodatabase in it. The long path can create problems especially if you run the tool via ArcPro. Make sure your environments are set up correctly to the acpf geodatabase where you store your outputs. When you run the tools check if the inputs and outputs have the correct path. Especially when you run the tool a multiple times like you do for the manual cutter.


Agnes Marchlewska
Project Manager
513-839-8272 (c)



David James

unread,
Mar 30, 2023, 3:17:55 PM3/30/23
to Agnieszka Marchlewska, Kurtis Fisher, ACPF Forum, Andy Asell
Kurt
I agree with Agnes regarding network drives... no network drives. My D:\ drive is a local drive, not a network drive. I usually keep my ACPF  toolbox and data on my D:\ drive and it works just fine. The documentation has a lot of good information on how to set up the ACPF workspace.

Dave James

McNeely, Robin L [DSN]

unread,
Mar 30, 2023, 3:36:28 PM3/30/23
to Kurtis Fisher, ACPF Forum, amarch...@gmail.com, ACPF Forum, Andy Asell

Kurtis,

 

I will address a few of your recent emails. First, everything Andy Asell has mentioned about the DEM being integer, centimeter Z is what we encourage from the ACPF Hub. If the DEM is currently floating point, you get the detail that the hillshade will show but more importantly, what the tools will use. However, that creates a very large file size and slows processing. Our suggested solution is to convert the DEM to integer but in order to keep the elevation detail, also convert the Z units to centimeter.

When you do this, you can effectively use a 1m horizontal unit which means processing at the HUC12 level is fairly quick and you get the detail the tools need to do an effective job.

As long as you save your data to a local drive on your computer, no matter the actual drive letter, your processing will go faster. Writing data, or reading data from, a network drive is not recommended.

As for your last comment about not using a watershed within the ACPF database, that is fine if you are creating your own but please follow the manual recommendations. The tools’ ability to run depend on expected naming syntax so please use what we set up. It may help you to download a fgdb in the core dataset to have an example.

If you are still having issues after all the feedback you get, please reach out to acpfs...@iastate.edu and one of the Tech team can work with you individually.

 

 

Text

Description automatically generated

Robin McNeely

GIS Technical Team Lead

mo...@iastate.edu  Iowa State University

National Hub for the Agricultural Conservation Planning Framework

https://acpf4watersheds.org/

 

cid:image002.png@01D81788.2D79D5A0

 

From: acpf-...@umn.edu <acpf-...@umn.edu> On Behalf Of Kurtis Fisher
Sent: Thursday, March 30, 2023 1:41 PM
To: ACPF Forum <acpf-...@umn.edu>
Cc: Kurtis Fisher <kjfi...@catamount.wcu.edu>; amarch...@gmail.com <amarch...@gmail.com>; ACPF Forum <acpf-...@umn.edu>; Andy Asell <andy....@dnr.iowa.gov>
Subject: Re: [ACPF Forum] Flow Paths Not Corrected by Cut Lines

 

I see now that both amarch...@gmail.com and Dave James pointed out my naming convention. Will this actually cause processing issues? I'm not working from watersheds within the ACPF database so I don't have a Huc12 folder to work from like in the videos.

On Thursday, March 30, 2023 at 1:35:36 PM UTC-5 Kurtis Fisher wrote:

My impression was that 1 meter or above is too detailed for watershed delineation, and I thought it would made sense to work within the same resolution as the one in which I will delineate my watersheds to. Degrading to 5 meters from 1 meter resolution was actually a method I used to enhance my 10 meter resolution. I have been warned that 1 meter resolution will be too fine-scaled and will cause issues in the watershed delineation process. Correct me if I'm wrong.

 

On Thursday, March 30, 2023 at 1:28:55 PM UTC-5 Kurtis Fisher wrote:

amarch...@gmail.com,

 

D drive is my internal solid state drive. Is this not what is meant by a local drive? Should this always be C Drive, no matter what?

On Thursday, March 30, 2023 at 12:49:00 PM UTC-5 amarch...@gmail.com wrote:

I’m not sure but maybe the name of your files might be an issue here. Also D drive is not recommended. You should try to run it on the C drive. I’m not sure what is the issue with your DEM but the hillshade looks off to me too. 

Sent from my iPhone



On Mar 30, 2023, at 1:32 PM, Kurtis Fisher <kjfi...@catamount.wcu.edu> wrote:

Hi Andy,

 

No, I didn't have a subset of lines selected. I actually made sure to clear all selected features at the end of each process for the sake of taking screen shots. 

 

I'm not sure why that would be the case for my terraced-looking hillshade... I'm using a 5 meter resolution DEM which I generated from a 1 meter DEM with a pretty straightforward cubic resampling method in ArcGIS. I just redid the DEM Preparation steps with an unaltered 10 meter DEM. It ended up looking the same way, just with less resolution. That's an interesting note about my straight-shot lines though, I didn't know it wasn't supposed to work that way... Both DEMS came straight from The National Map, 3DEP quality.

 

As far as the impeded flow, my cut lines did seem to make them disappear. So, my cut lines worked in a way, but I also need the flowlines to reflect these changes as I'll be using them to correct watershed boundaries in the catchment delineation process. Attached is the picture of my processing results.

 

Thanks,

 

-Kurtis Fisher

 

On Thursday, March 30, 2023 at 10:32:16 AM UTC-5 Andy Asell wrote:

Hmm, you didn't have a subset of your cutlines selected when you ran the Manual Cutter tool, did you? Another thing that looks kind of strange is your hillshade. It looks awfully "stair-steppy" (I'm having visions of rice paddies in China). It appears whatever algorithm used to generate the DEM/Hillshade didn't smooth out the elevations. Thus the reason you have a lot of "straight-shot" lines in your drainage network because they are running over areas of constant elevation. This would affect the viability of the TauDEM extensions when they're trying to model sheet flow. So the underlying original DEM might contribute to the wonkiness of some of your flowlines. But it doesn't answer your original question of why your cutter tool didn't take. Would you be able to reopen your last run of the Cutter too from the results tab and take a screenshot showing your inputs/outputs? Also, What does your updated "Identify Impeded Flow" based off your NewDEM look like after you've run the Manual Cutter? 

 

https://lh4.googleusercontent.com/LoT9fsTYWyDK0fa6XM0X_OHIOurSQFFbpglfxSgef4mpA20annose373fkrs-stfEYsMF-_FH8qWL7u-V36Zcl4zW8CxLA3m6j7ue2kiViK3Cq2bXkDCfa8RVwXOzPDK4COE7DdW

ANDY ASELL Environmental GIS Specialist

Water Quality Improvement Section

Iowa Department of Natural Resources

p 515-250-3303

502 E 9th St, Des Moines, IA 50319

https://lh3.googleusercontent.com/ebfRBiZq2I0qQQLj0tEwzIXXxjDNUHQTBPHSqjKluZT4t6foIMZMX3kouSFUiQvlwVWvWfeDWqTcAXAbi8kD4xnefxxSddyb7mjwwmqRnzL1IL0Fv0_Xf2X2VsXrjss1QwgHKxW-https://lh5.googleusercontent.com/sXZlaYsc_6gy6dDIM-GSDfu0V_uS4uZpwlUzo86Ez8iijVzYB85YEQ2b_pgt3ebx95kiP9KL9SQiG04Wab4lzHnRRA8rwROk2ux9nw2vnUiAJtGxvBT65JGJDDnnlLJ9yNgY9JCjhttps://lh3.googleusercontent.com/r-uy2EhAbk_BjEDd2UTRs4vrUK67bZXjwhsk8A_2Rbqf7QbIoN9TRqSCima-BQxvz7-rh6dhuhvjyCNZ8xz-Oi7di4luf_SA1-cMpu3V1XeelgmH00IeHVP4vcvhBBUCoSXfz2-Qhttps://lh4.googleusercontent.com/4mf0bWIJ1boOOvR-KWUSv2eramCsy_sKc-iiuy377TMDTKjVNyZiFsl_q-m9uWBmGRDorT0b0snMGTYHCqmSTAQYMkmkuhvW8lYOQ4atdZu1mtMS63cA3MPe7pddCjAtIxddd-3xhttps://lh3.googleusercontent.com/HcKDMLziExGFi6WrkzHLf_hFMaDyCixaRZ2t9FNLiOgXKRE42OHmcMBR5LHnmCcGtu3kE7b6cNhNhjZ5JAQWIUJPqDNdpOYyg59vAcnFkWaAv3BBYXix3j5R2dBvygc2k7-ZXvsx

 

 

Reply all
Reply to author
Forward
0 new messages