Get Discharge (frxst_pts_out) Output with Appropriate Forecast Point

[Cahyo]

Dear Mrs. Molly, Mrs. Katelyn, and All member of WRF Hydro Group. 

I am trying to run WRF-Hydro with WRF forcings and activate frxst_pts_out. 

How do you get the discharge output according to the forecast point that we specified in the GIS Preprocessor step? I did a scenario for the case of flooding in Jakarta, Indonesia (outside the U.S., river discharge observation data cannot be obtained from USGS) using the initial GFS 1 km data, 450 m hydrosheds. Expectations from the three forecast points that I specify the following discharge values:

(LON, LAT)

a. 106.84711 -6.205966

b. 106.831395 -6.400513

c. 106.835681 -6.630465

In reality, river discharge values actually appear at points that should not appear (LON, LAT): 107 degree, -6 degree.

How to fix it?

[Somaye Imani]

Dear Cahyo,

I also had faced this problem. To fix this problem, first, run the wrfhydro model based on forecast point and Fulldom_hires. Then open one of the outputs of Model named CHRTOUT in ArcGIS. You will see that forecast points are not located on the stream networks. So you should change the location of the forecast point on the stream network of CHRTOUT. Then you can use calculate geometry and extract new Lat and Long of forecast point on stream network and build Fulldom_hires again.

Sincerely,
Somaye Imani
Research fellow at Water Research Institute (WRI)

And

Ph.D. Candidate, Environmental Engineering, University of Tehran, Iran

Sender notified by Mailtrack 12/14/19, 08:56:10 AM

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/414853cd-7175-4270-a269-0dd9ad4af047%40ucar.edu.

[Cahyo]

How to open CHRTOUT Mr.Sony? Is it open directly or need to convert first ?

[Somaye Imani]

I convert hdf5 to Tif and after that, I opened in ArcGIS

Sincerely,
Somaye Imani
Research fellow at Water Research Institute (WRI)

And

Ph.D. Candidate, Environmental Engineering, University of Tehran, Iran

Sender notified by Mailtrack 12/14/19, 01:08:33 PM

On Sat, Dec 14, 2019 at 12:57 PM Cahyo <khol...@gmail.com> wrote:

How to open CHRTOUT Mr.Sony? Is it open directly or need to convert first ?

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/373f86dd-0c9c-448b-9f5f-9348b5dd7c8d%40ucar.edu.

[Cahyo]

So, CHRTOUT format is hdf5 ?
What an apps to use read and convert hdf5 to tif?

[Somaye Imani]

Yes all wrfhydro outputs are hdf5. There is no software for this but you can use ncdump command in linux cmd

On Sat, Dec 14, 2019 at 16:54 Cahyo <khol...@gmail.com> wrote:

So, CHRTOUT format is hdf5 ?
What an apps to use read and convert hdf5 to tif?

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/a331189b-0df2-4756-8983-f762760dc256%40ucar.edu.

--

[Austin Raney]

I would suggest looking into the Rwrfhydro package put together by the wrf-hydro team. They have a few vignette examples that should help in quickly extracting and analyzing output. 

Here is the link: 

https://ral.ucar.edu/projects/wrf_hydro/rwrfhydro

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/CAKoQsLhZ_n_LK5NnPrx%2BT9WXFnUn2nxwoxAOA%3DsbUt1ARRueoA%40mail.gmail.com.

--

Austin Raney
Undergraduate Research Assistant
Department of Geography
Surface Dynamics Modeling Lab 
The University of Alabama
2021 Shelby Hall
aar...@crimson.ua.edu | http://sdml.ua.edu

[Cahyo]

Thank for your suggestion, Mr. Austin.
I have installed rwrfhydro. However, many commands / vignettes that I have not understood. Which Vignette is used to fix my problem?

[Austin Raney]

Hey Cahyo,

This one should point you in the right direction. 

https://ral.ucar.edu/sites/default/files/public/streamflowEval_0.html

On Sun, Dec 15, 2019 at 23:18 Cahyo <khol...@gmail.com> wrote:

Thank for your suggestion, Mr. Austin.
I have installed rwrfhydro.  However, many commands / vignettes that I have not understood.  Which Vignette is used to fix my problem?

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/c8d69cc6-cadf-4b94-9410-1b01277ff365%40ucar.edu.

[Katelyn FitzGerald]

Thanks all for pointing Cahyo in the right direction.

Just to clarify here, most WRF-Hydro output files from recent versions are HDF5 files, but they also follow the netCDF-4 guidelines (in other words they are netCDF-4 files that are HDF5-based).  

There are a number of tools available to read and display netCDF files (there are also some for generic HDF5 files) some command line tools (e.g. NCO, CDO), some software libraries (e.g. R (including the rwrfhydro package mentioned in this thread), Python, Fortran, C), and some graphical tools (e.g. Panoply).  

Here's a more complete list from Unidata:

https://www.unidata.ucar.edu/software/netcdf/software.html

The Fulldom_hires.nc input file is also a netCDF file and you should be able to edit the 'frxst_pts’ variable with one of these tools or in GIS as you're creating the input files (the forecast points do need to be on the gridded channel though so you'll want to confirm that's the case when you move them).  As mentioned above, it sounds like the points you specified snapped to the wrong grid locations and just need to be corrected.     

At this point it's possible that not all of the input and output files have the required metadata to display appropriately in ArcGIS, so that might be the issue referenced above.  That said, Fulldom_hires.nc should be fine.

 

On Saturday, December 14, 2019 at 8:24:12 AM UTC-7, SONY wrote:

Yes all wrfhydro outputs are hdf5. There is no software for this but you can use ncdump command in linux cmd

On Sat, Dec 14, 2019 at 16:54 Cahyo wrote:

So, CHRTOUT format is hdf5 ?
What an apps to use read and convert hdf5 to tif?

--

Sincerely,
Somaye Imani

[Cahyo]

Dear, Mrs. Katelyn...

When I correct the forecast point using the arcgis preprocessing tool, do I just need to re-run the geogrid file? which option is best for the exact forecast point desired?

1. Check or uncheck the Mask Channel variable variable to forecast basin

2. Check or uncheck routelink

Maybe this option give an effect for output.

Thank you, Mrs. Katelyn.

[Kevin Sampson]

Cayo,

I suggest that you read the WRF-Hydro ArcGIS Pre-processing Tool documentation, section 9.1 "Manual Specification of Station Points", which describes the process of ensuring that your gage locations will be snapped to the correct channel pixel in your fulldom_hires.nc file. 

Here are answers to your questions:

1) You do not need to check the box "Mask CHANNELGRID variable to forecast basins?". This is an option if you would like to deactivate all channels do not contribute to your forecast points. That is up to you.

2) The "Create reach-based routing (RouteLink) files?" option is also not required and depends on the desired forecast configuration. If you wish to perform Muskingum or Muskingum-Cunge routing on a vector channel network, then select this option. If you wish to perform gridded diffusive-wave routing on your gridded channel network, then leave this option un-checked.

I suggest you read the Pre-procesor documentation and get familiar with these options. You can find the latest document here: https://github.com/NCAR/wrf_hydro_arcgis_preprocessor/releases/download/v5.1/WRFHydro_GIS_Preprocessor_Documentation_v5_1.pdf

In a previous email, you asked how to export Fulldom_hires.nc file variables to GeoTiff. The Fulldom_hires.nc file is CF-compliant and you can bring it into ArcGIS and QGIS a number of ways. In ArcMap, you can drag and drop the netCDF file into the map display or you can use the "Make NetCDF Raster Layer" tool in the Multidimension toolbox. See section 6.4.1 "Visualizing and Accessing NetCDF data in ArcGIS" for more information.

Thank you,

Kevin

Kevin Sampson, GISP
Associate Scientist
National Center for Atmospheric Research

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_users+unsubscribe@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/1a859e8c-9966-464f-b0b5-7b640baeae6d%40ucar.edu.

[Cahyo]

Mr. Kevin,
Thank you for your suggestion.
I have tried to do steps in section 9.1. However, the point is still wrong (or rounded behind the decimal point). I am using ArcGIS 10.6, does this version have an effect?

[Kevin Sampson]

Cahyo,

Could you provide some examples of what you mean when you say the point is still wrong? What is wrong about the point? Is it that your forecast point is not coinciding with the channel network in your Filldom_hires.nc file? If so, did you alter the coordinates of the gage in your CSV file to match the location on the gridded stream network where you expect the gage, and then re-ran the GIS pre-processor? Please provide images that describe the frxst_pts and channelgrid variables and the steps you have taken to resolve the issue.

Thanks,

Kevin

Kevin Sampson, GISP
Associate Scientist
National Center for Atmospheric Research
P.O. Box 3000 Boulder, CO  80307-3000
Tel: 303.497.8121

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/d750835c-8889-4738-a89c-11a4b7f066a7%40ucar.edu.

[Cahyo]

Mr. Kevin,

I attach a file containing the forecast point (.csv) and the output of the forecast point that was set after it was wrong (rerun GIS Preprocessing Tool with ArcGIS 10.6).

I set the point in the same position as 0 in channelgrid, but still no change. 

How to fix this setting ?

Pada Kamis, 19 Desember 2019 23.27.55 UTC+7, Kevin menulis:

Cahyo,

Could you provide some examples of what you mean when you say the point is still wrong? What is wrong about the point? Is it that your forecast point is not coinciding with the channel network in your Filldom_hires.nc file? If so, did you alter the coordinates of the gage in your CSV file to match the location on the gridded stream network where you expect the gage, and then re-ran the GIS pre-processor? Please provide images that describe the frxst_pts and channelgrid variables and the steps you have taken to resolve the issue.

Thanks,

Kevin

Kevin Sampson, GISP
Associate Scientist
National Center for Atmospheric Research
P.O. Box 3000 Boulder, CO  80307-3000
Tel: 303.497.8121

On Tue, Dec 17, 2019 at 9:44 PM Cahyo <khol...@gmail.com> wrote:

Mr. Kevin,
Thank you for your suggestion.
I have tried to do steps in section 9.1.  However, the point is still wrong (or rounded behind the decimal point).  I am using ArcGIS 10.6, does this version have an effect?

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hyd...@ucar.edu.

[Kevin Sampson]

Cayho,

I do not fully understand your question. Where are you looking that you think your forecast points are wrong? If you think the forecast points are in the wrong location on your channel network, then you can move them and re-run the GIS pre-processor as we described before. I had a look at the files and I see that the 'frxst_pts' variable in "Fulldom_hires.nc" places your forecast points on the channel network as defined by the 'CHANNELGRID' variable also in "Fulldom_hires.nc". I also plotted the coordinates from your "frxst-pts.csv" file and they are also specified on the channel network. Everything looks fine to me. I don't know where the 'frxst_pts_out.txt" file is coming from; is that being generated by WRF-Hydro? 

After looking at your files, I have some concerns about your domain. The domain is much larger than you seem to need. Also, it looks like the DEM you are using is much smaller than your domain. Please either change the size of your domain or use a DEM that has complete coverage over your domain, which is what we recommend in the GIS pre-processing documentation. Also, I would perform some treatment to your input DEM to make sure that ocean areas are specified as NoData in the input. Your channel networks also seem to end in the urban portion of your domain. Please review your DEM and the resulting routing grids to try to better understand and resolve the issues.

Good luck,

Kevin

Kevin Sampson, GISP
Associate Scientist
National Center for Atmospheric Research
P.O. Box 3000 Boulder, CO  80307-3000
Tel: 303.497.8121

To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/b071fd86-34cd-41ef-b15a-6e61f1710ff3%40ucar.edu.

[Cahyo]

Mr. Kevin,

Thank you for your suggestion, I will examine my DEM data or replace with hydrosheds.

Yes, frxst_pts_out.txt is being generated by wrf hydro which is from wrfout forcing.

I want to ask you, how to improve coefficient correlation value for my simulation in wrf hydro? Especially discharge output...

[Kevin Sampson]

Cahyo,

I believe that frxst_pts_out.txt is not a good place to look for forecast point coordinates. This file may truncate the precision of the coordinates and may not be very useful. A more accurate description of the location of your forecast points will be in the 'frxst_pts' variable in the 'Fulldom_hires.nc' routing grid file. 

We try to keep issues focused in both the wrf-hydro_users group and on the wrfhydro help email system. Please submit a separate request and fully describe your question regarding how to improve coefficient correlation values in your simulation.

Thank you,

Kevin

Kevin Sampson, GISP
Associate Scientist
National Center for Atmospheric Research
P.O. Box 3000 Boulder, CO  80307-3000
Tel: 303.497.8121

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/92bb9bf3-9cf9-44d3-bf09-80693318ccc2%40ucar.edu.