Downscale Function created for for hourly data in nc

[Snafkin547]

Description

This new function, called 'NA_downscale_hrly', is extended from the original 'NA_downscale' function.
This newly created downscale function handles hourly data stored in nc file and generates outputs including the weights of each ensemble member for each observatory site. These weights are not currently used, however, they will be used effectively in the aggregation step in the future, once we come up with an idea of how to distribute to each coordinate.

Motivation and Context

In the previously developed 'NA_downscale' function, monthly-yearly data was handled. This newly developed downscale function enables to deal with the hourly data. Due to the difference in data structures and absence of weights in previous data structure, we needed this change.

Review Time Estimate

• Immediately
• Within one week
• When possible

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My change requires a change to the documentation.
• My name is in the list of CITATION.cff
• I have updated the CHANGELOG.md.
• I have updated the documentation accordingly.
• I have read the CONTRIBUTING document.
• I have added tests to cover my changes.
• All new and existing tests passed.

[Snafkin547]

Example usage

# Set working directory
my_wd <- "<Your working directory>"
setwd(my_wd)

# Load necessary library
library(terra)
library(readr)
library(sf)
library(ncdf4)

nc_file <- "/Users/hishii/Desktop/PEcAn/osfstorage-archive/NCDF/NEE.nc"
nc_data <- nc_open(nc_file)
coords <- paste0(my_wd, "data/SDA/sites.csv")
covariates <- rast(paste0(my_wd, "data/covariates.tif"))
date <- "1986-01-01 00:00:00 EST"
downscale_output <- NA_downscale_hrly(nc_data, coords, date, covariates)

[mdietze]

Overall, function is a good first start but doesn't really address the need to downscale timeseries, but instead provides a way to downscale a single time point. I'm happy to pull in this PR as an initial milestone without that generalization once the other smaller points are addressed.

[mdietze]

Doesn't need any change for this PR, but making a note for those doing future development that this would be a great place to either parallelize this work or to modify the ML model to take time into account explicitly.

[infotroph]

Hi, @Snafkin547! With this PR you became an author of the PEcAn.assim.sequential package, so we'd like to make sure we have your agreement to change the package license from NCSA to BSD-3. Commenting "yes" here is enough. More details are in #3364 if you want them.

We started the relicensing process (with notes in CHANGELOG.md) before you opened this PR, but at merge time we hadn't yet updated any license files in this package, so for full clarity we'd like to hear your explicit consent even if you had been assuming your contribution would be BSD-3 all along 😁