Fetch a single year of daily BioSIM Fire Weather Index for one location batch

Wraps BioSIM::generateWeather() for the FWI_Daily model, fetching one calendar year for the cells in locations_batch[[1]]. Output is appended to an Arrow CSV dataset under path/FWI_Daily/, partitioned by YEAR/BatchID. Like get_clim_daily(), 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_fwi_daily(
  locations_batch,
  year,
  studyArea_hash,
  params = NULL,
  path = .climateCachePath()
)

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.
params: list of additional parameters forwarded to BioSIM as additionalParms = list(FWI_Daily = params). Default NULL.
path: character. Directory under which the FWI_Daily/ arrow dataset is written. Default uses the package climate cache (getOption("landisutils.cache.path")).

Value

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

Details

BioSIM's FWI_Daily model is known to occasionally produce implausibly large FFMC, ISI, and FWI values (see https://github.com/RNCan/BioSimClient_R/issues/14). To work around this, FFMC values \(>\) 101 are recomputed from the underlying T/RH/WS/Prcp columns using cffdrs, and ISI and FWI are then recomputed from the corrected FFMC. The unused DSR column is dropped.

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. A small random delay (5-30 s) is applied per call to spread BioSIM hits when many parallel workers start at once.

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