Welcome to the Spring 2025 class forum!
Please use this topic to post any message you want related to the Spring 2025 short course!
1 Like
As a reminder, the first homework assignment is for you to get familiar with the VICTOR platform. Please copy and run one notebook of the 2023 or 2024 short courses (located in the /shared folder) or install one of the models (using the âvictor setupâ command in the terminal) and report any problems and questions using this forum!
2 Likes
Thanks for the information, I will get familiar with the platform.
I tried running ols_mogi_example.ipynb from Part 3 of the Spring 2023 course, and installing the Hysplit model using âvictor setupâ. Both went smoothly for me. Thanks!
I copied the dome radius hazard model in the Spring 2023 into my home directory and ran it with no problems.
I tried installing Conflow via âvictor setupâ command in the terminal and the folder of Conflow appeared in my folder tree. I tried running each cell and played around with the choice of outputs. I intended to download/save the output figure (if itâs allowed/possible), I still donât know how to, but I guess I need to explore further the documentation/tutorial/follow the next meetings.
I was able to run the Spring 2023 âadvection-diffusionâ notebook. Most of it looks similar to code I have seen from Chuck previously, so I was able to compare and make sure it worked properly.
Hi! I ran ShortCourse.ipynb yesterday and came across this curiosity. I wanted to create DEMs for Awu, a volcano off the north coast of Sulawesi (Indonesia). Running line [2] of the notebook (#Choose a volcano and DEM options) with the name âAwuâ generates a DEM with centred latitude and longitude of (-5.05, 151.33), which actually corresponds to Ulawun volcano in Papua New Guinea. You can see it in this useful coordinate converter and visualiser: Lat/Lon and UTM Conversion - Yellowstone Research Coordination Network. This is likely something to do with the input volcano name (Awu vs Ulawun) and might be useful to check for others creating DEMs for volcanoes which have similar names (e.g., Soufrière, Cerro Azul, Rumble).
Hi Ailsa,
Thanks for sharing this. Yes, the search by name has limitations, and may not be the best option for volcanoes that are less known and/or that are sharing the same few letters as better known ones. Your example clearly shows this and in such cases, using the interactive map on OpenTopography as we showed yesterday in class may be a better option. In any case, please always double check the coordinate values you get to make sure the DEM is the one you want! 
1 Like
For the section of the code where the line volcano = victor.search_volcano()⌠is defined, when we run it, we get an error with the victor library, stating that victor is not defined.
we tried installing victor directly and then importing the library, but it doesnât work.
Also, the code in Victor indicates that it depends on a local Excel file for the search_volcano() process, which makes it difficult to connect with the Smithsonian database. The Excel file is named GVP_Volcano_List_Holocene.xlsx.
How can we fix this error?
Hi Ailsa,
The search function returns a list of all volcanoes with the included input (in your case Ulawun, Lawu, Mahawu, and Awu). If you would like additional functionality (printing matches, etc), please let me know!
1 Like
Hi Paulo,
You must run the notebook from the beginning, as without the cell with the imports, the code will not know what functions to reference.
The code does depend on a local Excel file, which is uniformly formatted and was provided to us by Sarah Ogburn. It is not meant to be a consistent copy of the Smithsonian database. If you have changes/recommendations, we would appreciate any feedback!
Hi, Iâm struggling to get my DEM to reproject to UTM. When I check the crs, it says EPSG. I think this is resulting in subsequent steps having only null values in the data array, and so canât create an elevation profile. Has anyone else had this problem? Any thoughts on how I could resolve this?
Hi, I also have the same problem with Ailsa as shown by the following screenshot:
In case I still want to continue running in the notebook, how do I choose the second column for Marapi volcano that I am interested in?
Itâs not a perfect solution, but I found a workaround based on this StackOverflow thread: How do I check for an exact word or phrase in a string in Python - Stack Overflow.
Basically, adding " r'\b[Volcano name]\b' around your desired volcano name will search the GVP database for that exact name (or string), rather than all database entries containing those letters.
First pic - using 'search_volcano' function with 'Awu':
Second pic - using
'search_volcano' function with r'\bAwu\b':
(The second time, I also used
search_volcano? to tell me what the function does.)So for your solution @dnurfiani, try using:
victor.search_volcano(r'\bMarapi\b')
Thank you Ailsa, it works.
1 Like
Iâve got stuck on extracting a profile of my chosen DEM. I followed the solution notes, slicing my DEM first by X (slice 1000 rows) then by Y (slice 1 column):
demarray_cut = demarray.sel(x=slice(utm_east, utm_east+1000))
demarray_cut = demarray_cut.sel(y=utm_north, method="nearest")
Which produces this slice with array values all = 0.0:
This produces a very disappointing profile
-
Interestingly, if I slice only by x-coordinates this line of code
demarray_cut = demarray.sel(x=slice(utm_east, utm_east+50)) I produce an apparently realistic profile of Tungurahua volcano (may be baloney, though):
I think Iâm confusing the array coordinates (X rows x Y columns in which data values are stored) with array data (Z-values representing elevation), but canât figure out how to slice by coordinates and then plot elevation data. Help appreciated!
Hi! Iâm also having issues with searching for my volcano. When I search for âHoodâ for Mt Hood, it seems to bring up the wrong coordinates and DEM, despite GVP calling it by this name iâm entering. I was wondering if there was a way I can manually add the correct lat and long coords to bring up the Mt Hood DEM?
Hi Alexandra,
You can manually set the lat/lon with the corresponding variables in the notebook. Alternatively, you can look inside the volcano variable and select a different option. With âHoodâ, Hoodoo Mountain matches first, but both that and Mt. Hood are returned in the variable. Instead of lat = volcano[1][0], it would be lat = volcano[1][1] in this case. Iâll work on generalizing the notebook to make this process easier in the future.
I tried entering the lat/lon into those appropriate lines but it looks like it is still displaying Hoodoo, I may have entered it in the wrong place! I also tried changing it to lat = volcano[1][1] but no change by looks of it, can you spot where I went wrong in this screenshot/how to just pick out the Hood coords rather than Hoodoo as its trying to do?
