# Creating a vertical grid cross section in Wradlib **URL:** \ **Category:** General\ **Created:** [May 3, 2023, 6:00am UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232 "2023-05-03T06:00:57Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![adithiyrnair](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/adithiyrnair/32/84_2.png) [@adithiyrnair](https://openradar.discourse.group/u/adithiyrnair)\ **Post date:** [May 3, 2023, 6:00am UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/1 "2023-05-03T06:00:57Z")
I produced a gridded product of Reflectivity using wradlib and I want to take a vertical grid cross-section with a start and end point in terms of lat and long. How to do it ? ---
**Author:** ![mgrover1](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/mgrover1/32/7_2.png) [@mgrover1](https://openradar.discourse.group/u/mgrover1)\ **Post date:** [May 3, 2023, 12:48pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/2 "2023-05-03T12:48:06Z")
If it is an `xarray.Dataset`, I would use metpy’s cross section functionality! [https://unidata.github.io/MetPy/latest/examples/cross\_section.html](https://unidata.github.io/MetPy/latest/examples/cross_section.html) This is what is used in Py-ART under the hood for the `plot_cross_section` method of the `GridDisplay` ---
**Author:** ![adithiyrnair](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/adithiyrnair/32/84_2.png) [@adithiyrnair](https://openradar.discourse.group/u/adithiyrnair)\ **Post date:** [May 3, 2023, 12:55pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/3 "2023-05-03T12:55:01Z")
@mgrover1 Thanks for your reply. Let me check it out this metpy fuction I used this Py-ART plot\_cross\_section. The problem is the interpolation (Barnes, Barnes2, Cressman and Nearest). All interpolation gives me unwanted results, especially near cloud edges. But the interpolation technique used by wradlib gave me better results. So I thought is there any technique to plot vertical cross sections just like Py-ART is available on Wradlib? ---
**Author:** ![mgrover1](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/mgrover1/32/7_2.png) [@mgrover1](https://openradar.discourse.group/u/mgrover1)\ **Post date:** [May 3, 2023, 1:00pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/4 "2023-05-03T13:00:40Z")
Using the metpy functionality here would give you the same result as using the plot\_cross\_section function. It works generically on xarray datasets, with the exact line here: > You should be able to apply this to grid output from wradlib, as long as it is an xarray dataset. ---
**Author:** ![adithiyrnair](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/adithiyrnair/32/84_2.png) [@adithiyrnair](https://openradar.discourse.group/u/adithiyrnair)\ **Post date:** [May 3, 2023, 1:15pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/5 "2023-05-03T13:15:10Z")
@mgrover1 This is applicable for pyart grid object which saved as NC file right ? ---
**Author:** ![mgrover1](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/mgrover1/32/7_2.png) [@mgrover1](https://openradar.discourse.group/u/mgrover1)\ **Post date:** [May 3, 2023, 1:24pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/6 "2023-05-03T13:24:07Z")
@adithiyrnair yup! You can read that back in using xarray since it is a netcdf file. ---
**Author:** ![scollis](https://yyz2.discourse-cdn.com/free1/user_avatar/openradar.discourse.group/scollis/32/21_2.png) [@scollis](https://openradar.discourse.group/u/scollis)\ **Post date:** [May 3, 2023, 2:15pm UTC](https://openradar.discourse.group/t/creating-a-vertical-grid-cross-section-in-wradlib/232/7 "2023-05-03T14:15:32Z")
@adithiyrnair Thats where you need to fill in fill value areas with a number. Some times I have gone back in and calculated what the minimum detectable signal will be and filled in fill values with that to give the algorithm something to grid to. Otherwise if convolutes (smears) the value out to the search radius.