Pour Point for Watershed Analysis?

[WilliamPalmquist]

This is going to seem like a very elementary question, but I have not been able to find information describing what a 'pour point' is, in regards to conducting a watershed analysis.

I have the flow_direction, flow_accumulation, and fill layers all set, but in the Watershed analysis toolbox, it asks for the 'pour point' either in raster or vector form (presumably after being run through the 'snap pour point' tool).

How do I find/make pour points in order to complete the Watershed analysis? Also, what is the best way to use 'snap pour point' to ensure I have the correct locations?

Thanks very much,

Will Palmquist

[CaitlanReilly]

Will,
I haven't done watershed stuff in a while, but I think the "pour point" is the outlet of the watershed.  I'm not 100% sure, but I think you can define the outlet as a point from another shapefile, or perhaps a point from a raster.  I also thought you also had the option to define a point sort of manually, but I could be wrong. 

Hope this helps,
Caitlan

[CaitlanReilly]

Also I think that the snap pour points tool will assign pour point(s) to the pixels with the highest values of accumulated flow, according to the tool help.

[LornaMurison]

Caitlan is right, it is the outlet of the watershed.  All water within the watershed should drain to that one point.
Where you get the pour point from will depend on what the purpose of your project is.
The "snap pour point" tool is used after you have the original pour points.  It will take each point and snap it to the cell with the highest flow accumulation within a specified distance. That distance will depend on the resolution of your flow accumulation grid, and on how accurate you believe your pour points to be in the first place.  It basically makes sure that the point actually lands on a river and not right next to it.

Hope this helps

[MichaelBrady]

Hi,

This comment is in response to the statement made in the last post: "Where you get the pour point from will depend on what the purpose of your project is." I'm developing a model to estimate discharge into a River. I haven't been unable to find guidance on where to indicate outlets to create "projects" in geoHMS. Can anyone point me to any documents/discussions that outline things to consider when determining outlet points for a project with the aim to estimate tributary flow into a river? This seems to me like a critical step in the process and I've found the ACoE geoHMS tutorial to not be very thorough in explanation with this step.

My initial idea is to pick locations with USGS stations, but I have hundreds to choose from, so I'm not sure which to pick. Also, I would have to move the stations in some cases hundreds of meters to land on one of the synthetic stream pixels, which is another issue I'm not sure about, but I'm hoping it will make more sense as I continue with the model setup.

Any guidance will be more than welcome.

Thanks!

[toddsams]

I have not worked with geoHMS. However, it seems that selecting existing USGS gages will be helpful, mainly because you have measured flow data at those locations. You will have to move the points (as you stated) to match the synthetic stream grid. I would suggest working from a copy of your original points. This can be a bit tricky, especially when the gage is located near a stream junction. However, the USGS gages should generally have reliable lat/long data and they each will typically have a good location description (stream name etc.) associated with them. They should also have a pre-determined watershed area attribute which you can use to compare with the watershed areas from the polygons you generate. You can then go back and fix the outliers, if any.

Another tip is to have the digital USGS topo sheets and High Res NHD available to draw in the background. These will provide perspective when determining which stream cell the gage should match up with.

I will say that the whole process is a somewhat tedious task. However, it is important to do this manually. Using an automated "snapping" procedure will almost always lead to improper watershed delineations for at least a few sites (if not many).

[MichaelBrady]

Hi Todd,

Your advice was very helpful. I'm about 90% done creating subbasins for the USGS gauges that were in service during the rain events I plan to study.

My goal is to simulate discharge into the Hudson River in NY. I was surprised to find few gauges at streams entering the river below Troy Dam in service during the past 30 years. This is my first time setting up an HMS model, so I'm a bit unclear as to how the project points will be used (e.g., will data output at gauge stations be used primarily to calibrate the simulation model and should other drainage points of interest - i.e., ends of streams with no gauges just before they enter the river, be indicated during the data model setup, etc.?).

My current plan is to create additional subbasins that drain into a stream cell just before the Hudson (location with no USGS gauge stations). Does anyone know if this is common practice for estimating discharge into a large river? Is there a better method?

-Mike