Fetch a single year of daily BioSIM weather for one location batch

Wraps BioSIM::generateWeather() for the ClimaticEx_Daily model, fetching one calendar year for the cells in locations_batch[[1]]. Output is appended to an Arrow CSV dataset under path/ClimaticEx_Daily/<studyArea_hash>/<rcp>_<clim_model>/, partitioned by YEAR/BatchID. The function is intended to be run as a dynamic target across many (batch, year) combinations; existing partition files are detected via file.exists() and skipped, so re-runs are cheap.

Usage

get_clim_daily(
  locations_batch,
  year,
  studyArea_hash,
  path = .climateCachePath(),
  rcp = "RCP45",
  clim_model = "RCM4"
)

Arguments

locations_batch: list whose first element is a data frame with columns ID, Latitude, Longitude, Elevation, EcoID, BatchID. Typically one element of create_locations_df()'s output, wrapped in a list.
year: integer single calendar year to fetch.
studyArea_hash: character. Short hash of the study-area object (see .studyArea_hash()) used as a cache subdirectory so that distinct study areas don't collide.
path: character. Directory under which the ClimaticEx_Daily/ arrow dataset is written. Default uses the package climate cache (getOption("landisutils.cache.path")).
rcp: character. BioSIM representative concentration pathway. One of "CONSTANT_CLIMATE", "RCP45", "RCP85". Default "RCP45" (BioSIM's own default).
clim_model: character. BioSIM climate model. One of "RCM4", "GCM4", "Hadley". Default "RCM4" (BioSIM's own default).

Value

Path (character) to the partition CSV for (year, BatchID).

Details

For years within BioSIM's observational range, rcp and clim_model have no effect on the returned data, but they do still namespace the cache. For future years they select the CMIP5-era projection: rcp is one of "CONSTANT_CLIMATE", "RCP45", or "RCP85"; clim_model is one of "RCM4" (Canadian Regional Climate Model 4, CanESM2-driven), "GCM4" (CGCM4 / CanESM2-family GCM), or "Hadley" (HadGEM).

Rate-limiting against the BioSIM web service is the caller's responsibility - cap parallel workers (e.g. future::plan(multisession, workers = 4)) or stagger orchestrator calls. This function performs no internal throttling.

Topographic considerations

The climate data sources used by this package (BioSIM, Daymet, TerraClim) are not PRISM-derived. PRISM (https://prism.oregonstate.edu/) applies topographically-aware interpolation (slope, aspect, elevation, coastal proximity, temperature inversions) when downscaling station observations to a grid (Daly et al. 2008); the sources wrapped here use simpler interpolation schemes that can produce large discrepancies in areas of complex topography (mountainous terrain, steep elevation gradients, rain shadows): inter-product differences of 5-60% in annual precipitation (Henn et al. 2018) and >6 \(^\circ\)C in temperature (Walton & Hall 2018) have been documented in the western US. Carefully evaluate the suitability of the chosen source for your study area, especially in topographically heterogeneous landscapes.

References

Daly, C., Halbleib, M., Smith, J.I., Gibson, W.P., Doggett, M.K., Taylor, G.H., Curtis, J., & Pasteris, P.P. (2008). Physiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States. International Journal of Climatology, 28(15), 2031-2064. doi:10.1002/joc.1688

Henn, B., Newman, A.J., Livneh, B., Daly, C., & Lundquist, J.D. (2018). An assessment of differences in gridded precipitation datasets in complex terrain. Journal of Hydrology, 556, 1205-1219. doi:10.1016/j.jhydrol.2017.03.008