ornldaac / daymet-single-pixel-batch

Automating the download of multiple locations for the Daymet Single Pixel Tool

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

(When) will the monthly latency data be available through the single pixel API?

dlebauer opened this issue · comments

similar to #2 but ... wondering if / when the monthly latency data available at https://thredds.daac.ornl.gov/thredds/catalog/ornldaac/1904/catalog.html will be made through the single pixel API?

This came up in a question about accessing monthly latency data through the daymetr package bluegreen-labs/daymetr#50

Also, thanks for this awesome data product!

And if it is helpful, I bumped the end year from 2020 to 2021 in master...dlebauer:update_end_year_2021 (but I also realized this is probably too easy - that it might be necessary to update the code to handle partial years).

@dlebauer We have that in our sites, but it requires a couple things... 1. the files accessed by the SPET tool are optimized for speedy extraction. We are currently not producing that optimized format in our workflow. And yes, the code needs to updated to handle monthly files and partial years. Also, importantly, because of the spin up required to run a month of data, a variable/ or perhaps some of the variables/ don't mix with the data produced on the annual timing. In particular, Snow Water Equivalent. Other variables are likely fine, but we need to do some analysis to be sure.

Hi @michele-personal thank you for your reply! When you say 'we have that in our sites' do you mean 'planned, but not on the schedule'? Either way, there are a lot of people who would be interested in such a product!

In the short term, is there a recommended way to convert the tiles into files that can be queried by lat/lon? From the projected netcdf grids that are available, it appears like a fairly complex set of steps to get from lat/lon to x/y that someone must've already figured out - either built into a tool like gdal or part of your workflow that we could use offline?

I think within our resources, the best method is to extract from the data available through OPeNDAP. You'll need to convert your lon,lat to projected x,y. There are a number of ways to do that. Let me know if you need additional information for the projection and/or OPeNDAP access.
This Resource may be helpful.
https://github.com/ornldaac/thredds_opendap_r_max_temperature

I'm currently in the same situation it looks like @dlebauer was in a few months ago. I'm trying to extract information from the latency .nc files for a set of lat/lons in R, but I a) can't figure out how to easily convert to projected x,y and b) can't find any other information online. @michele-personal the resource you linked to when you resolved the issue is no longer active - do either of you have any resource recommendations? Thanks so much :)

@ansbros We removed that tutorial as it's becoming deprecated, but I've attached the file here for your convenience.
opendap_r_v1.zip
You might also find helpful code snippets here: https://github.com/ornldaac/web_services_data_access
Or there are other Learning Resources here: https://daac.ornl.gov/resources/learning/

In addition, the proj4 utility is often a convenient method to convert from lat/lon to Daymet's crs.
Hope this helps!

Also @ansbros The 2021 daily data are now available through the Single Pixel Extraction Tool. If that's helpful.