My flowlines all go uphill, why?

3358
24
Jump to solution
10-27-2016 01:59 PM
HillaryHudson
New Contributor II

I have created a set of flowlines using the Fill, Flow Direction, Flow Accumulation, and Reclass tools (in that order) from the spatial analyst toolbox in 10.3.1. I then converted them to polylines from the original raster data. When I looked at my flow directions in the vector data, they were all going uphill. I would like some help trying to figure out what I did wrong. Thank you.

0 Kudos
24 Replies
HillaryHudson
New Contributor II

Dan Patterson, I just took note of the example of they used in the flip tool link you sent. It would appear that the flow direction tool may be wonky when you apply it to stream data, from what that example shows. From the feedback I am getting,  it seems that the reason behind why the tool is not working properly isn't a known quantity, but flipping the lines is. Sigh. I wish I could get some answer regarding why the tool is having problems from the ESRI developers rather than just figuring out how to fix the bizarre output.

0 Kudos
HillaryHudson
New Contributor II

Thanks Darren for the script on for the batch flip. It looks like that is going to be the solution that I use.

0 Kudos
by Anonymous User
Not applicable

Hillary Hudson
Hi Hillary, 

I can take a look at this and see if it is an issue.
Could you please zip the elevation data you are using and send that along with a brief description of your workflow to NRajasekar@esri.com ?

Thanks,
Neeraj 

HillaryHudson
New Contributor II

Ok, I will do that. I am compressing the elevation data now, though I fear it will be too large to email. It is an entire watershed of LiDAR data.

0 Kudos
by Anonymous User
Not applicable

Would you know if this data is publicly available? 

In one of your comments you mentioned, "My stream orders all made sense and seem to follow the correct flow direction, the flow direction output itself, however, is backward." Could you please clip the flow direction output raster & input elevation raster, for a region where you can see that it is backwards? If you could send that, it might be sufficient enough for me to reproduce what you are seeing. In which case, you wouldn't need to zip the whole watershed elevation data and send that.

In the description of your workflow (Fill, Flow Direction, Flow Accumulation, and Reclass tools as you've indicated in your question) , please remember to mention what parameters options you use. If there are specific environments (other than defaults) you set for any tool, mentioning those would be useful as well.

If I can reproduce the issue with this data & information, we can look into it further .

HillaryHudson
New Contributor II

No, it’s not. It was acquired by Santa Fe County, NM and I do not believe they post their data to the web.

0 Kudos
HillaryHudson
New Contributor II

I am resending my data as two emails.

0 Kudos
NeilAyres
MVP Alum

You should actually just zip the entire shapefile into one zip. Downloading each individual component separately will be a bit of a pain.

An image of the dem generated from the lidar would also help a lot.

HillaryHudson
New Contributor II

Really? In Outlook you can select, copy, and paste all of the file components to your save location and that takes less effort than compressing them.

I whether it is a DEM and/or the LiDAR data I still need a location to post the data to since it is too large to send as an attachment.

0 Kudos
NeilAyres
MVP Alum

It's just a small piece of it. You don't have to post everything.

And in your previous message there were a total of 8 separate zip files. I don't look at this stuff using outlook.

How about an image of a piece of your dem and the flow direction derived raster. That would be a start.

Because there is obviously something wrong with your input data. Many, many people have successfully derived stream lines from dems. And they don't go uphill.

0 Kudos