No sediment or water output

195 views
Skip to first unread message

Bastian Ringleb

unread,
Oct 28, 2020, 9:14:58 AM10/28/20
to caesar-lisflood

Dear all,

I am new to LEMs and for my master thesis I am trying to model sediment and water output from a catchment in the Israelian negev.

The catchment has an area of ~2.2 ha, is located on a slope, and ends in a cistern. The original DEM has a pixelresolution of 7cm. I changed that to 49cm.

For my setup I edited the DEM. I filled all the sinks and edited the walls (boundaries) of the catchment to a minimum of 2 pixels wide (so no water can leave the catchment). I filled the cistern (end point of the catchment) to be level with the surrounding terrain. The ArcGISs Flowaccumulation tool now shows a continuous stream to the end point.

I rotated it to 90 degrees and added a channel to reach the right-hand side of the DEM. The DEM now consists of ~92k pixel.

To test the DEM in catchment mode, I tried different rainfall data. I used precipitation data from a weather station nearby (very little rain). I also tried a heavy edited rainfall file (very high precipitation).

For all the different setups I used the model looks like it is not running, but shows the processed time at the bottom (slow for the first 2 days, then a very fast processing speed). For all this time I am not getting an output for either sediment or water (Qs; Qw). When I change the view to the erosion tab it shows a little change, which suggests that something is happening.

Is the resolution to low or is the exit point at a wrong position? I attached some pictures of my model and hope someone can help.

 

Best regards

Bastian

Flow_acc.png
DEM.png
DEM_channel_90d.png

Nadav Peleg

unread,
Oct 28, 2020, 10:01:22 AM10/28/20
to caesar-lisflood
Dear Bastian,

I am not an expert to the model but know the study area a bit. The model is slow for the first two days as it computes the water flow, then likely you do not have any flow computed and the model can run faster. I assume you the first day is wet? What is the time step of the rainfall input you are using? Also, it can be that the flow is not reaching the outlet you created at the east - check the flow depth maps and see if you have flow somewhere in your domain right after the peak rainfall.

Best,
Nadav

Bastian Ringleb

unread,
Oct 28, 2020, 10:19:28 AM10/28/20
to caesar-lisflood

Dear Nadav,

 

Thank you for your quick answer. This is the same I assumed. So, I edited my rain file, that there is major rainfall after 20/50 days. The first two days have rainfall but not significantly more than the other days. I even tried the rain-file provided from the quick start guide. I use 60min timesteps.

I forgot to mention earlier that the flow depth maps don’t show anything. So, I think that it is not processing the rainfall data. I attached a picture of the drainage area, maybe you can spot an error there.

Best,

Bastian

drainage.png

tom.co...@gmail.com

unread,
Oct 28, 2020, 10:29:33 AM10/28/20
to caesar-...@googlegroups.com

Hi Bastian & Nadav,

I was going to suggest the outlet issue – as it doesn’t look quite like your channel/canal reaches the edge of the DEM – though it may just be how its clipped in the image. Is your rainfall rate in mm/hour? This is important…

Tom

 

PS – no need to rotate the DEM for the last couple of years! (sorry!)

--
You received this message because you are subscribed to the Google Groups "caesar-lisflood" group.
To unsubscribe from this group and stop receiving emails from it, send an email to caesar-lisflo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/caesar-lisflood/619484c9-7d21-450d-a35b-9f6791b44985n%40googlegroups.com.

Bastian Ringleb

unread,
Oct 28, 2020, 10:57:21 AM10/28/20
to caesar-lisflood

Hi Tom,

I thought the same and checked the outlet several times. In RasterEdit the canals right-most pixels are filed. I attached it, if you want to take a further look.

 

The first rainfall data I used had a 10min resolution. I added the values together in chunks of 6 to get the mm/hour rate. I even tried the rainfall file from the quick start guide (swale) as this quick start model worked fine for me. But for both runs I am not getting the flow depth calculated.

 PS – no need to rotate the DEM for the last couple of years! (sorry!)“ – Can I keep the orientation and as long an outlet reaches an edge the program computes the output?

Best regards,

Bastian

DEM.txt

Tom Coulthard

unread,
Oct 28, 2020, 11:04:40 AM10/28/20
to caesar-...@googlegroups.com
Yes to the rotation point. You’ll need your canal to the edge still. 

Ok. Try editing rainfall (increase amounts). 

Otherwise - what is your m value, MinQ value and DEM resolution 

On 28 Oct 2020, at 14:57, 'Bastian Ringleb' via caesar-lisflood <caesar-...@googlegroups.com> wrote:



Bastian Ringleb

unread,
Oct 28, 2020, 11:23:00 AM10/28/20
to caesar-lisflood

The DEM has a resolution of 49cm (the original data was 7cm). Therefor I choose a minQ of 0.0049 (resolution/100). I also tried higher values, resulting in a variable check error.

For m I chose 0.005, because of the arid environment of my study area. I also tried different values here, to see if I can get different results.

Bastian Ringleb

unread,
Oct 29, 2020, 6:23:52 AM10/29/20
to caesar-lisflood

I found the issue. When I changed the resolution to 0.49 m, ArcGis was showing exactly this resolution but the header of the rasterfile showed a resolution of 0.48999999. After I changed it in the header to 0.49, the model was running fine.

Thanks for all the answers!

Tom Coulthard

unread,
Oct 29, 2020, 7:16:46 AM10/29/20
to caesar-...@googlegroups.com
👍👍

On 29 Oct 2020, at 10:23, 'Bastian Ringleb' via caesar-lisflood <caesar-...@googlegroups.com> wrote:


Reply all
Reply to author
Forward
0 new messages