rainOrSnowTools

The goal of the rainOrSnowTools R package is to support analysis for the Mountain Rain or Snow citizen science project.

🌨️🌧️👉 The processed data are available on our public-facing dashboard 👈🌧️🌨️

You can learn more about how to use the data on the dashboard’s User Guide tab.

---

rainOrSnowTools provides:

1. Access to meteorological data from the HADS, LCD, and WCC networks.
2. Modeled meteorological data (air/dew point/wet bulb temperature and relative humidity) for an observation point.
3. GPM IMERG probability of liquid precipitation (PLP) data for an observation point.
4. Geographical data (elevation, state, and Ecoregion 3/4) for an observation point.
5. QC of processed observation data.

---

Installation and Loading

You can install rainOrSnowTools with:

# install.packages("devtools")
devtools::install_github("LynkerIntel/rainOrSnowTools")

Load in the rainOrSnowTools library:

library(rainOrSnowTools)

Package Functions

Each observation is geotagged with a datetime, location (latitude, longitude) and phase observation.

In this example, we will use the following sample: datetime: 2025-04-10 12:30:00 UTC,lat = 40,lon = -105, phase = "Snow"

Tagging an observation with ancillary data:

Geography-related:

Need to provide a lat/lon

Function	Description
get_elev	Elevation in meters, derived from the USGS 3DEP 10m product
get_eco_level3	EPA Ecoregion Level 3
get_eco_level4	EPA Ecoregion Level 4
get_state	U.S. State

# Elevation data in meters:
elev <- rainOrSnowTools::get_elev(lon_obs = lon,
                                  lat_obs = lat)
# [1] 1590.199

Meteorology-related:

Need to provide datetime and lat/lon

Function	Description
get_imerg	GPM IMERG Probability of Liquid Precipitation value (closer to 0 = likely snow, closer to 100 = likely rain)
model_meteo	This final function uses data collected from helper functions to model met variables for an observation point. Will only provide modeled data for points with >= 5 station data. Helper functions = access_meteo, qc_meteo, select_meteo, gather_meta

• access_meteo = Access data (+/- 1 hour) from the met network of choice (HADS, LCD, WCC, or ALL). Required variables include datetime, lat, lon, deg_filter (search radius)
• qc_meteo = QC the met data collected using standard thresholds
• select_meteo = Final selection of the met data from the network(s) of choice, filtering data to grab the closest in time
• gather_meta = Gathers the metadata associated for each station from select_meteo

# Get GPM IMERG probability of liquid precipitation:
plp <- rainOrSnowTools::get_imerg(datetime_utc    = datetime,
                                  lon_obs         = lon,
                                  lat_obs         = lat,
                                  product_version = 'GPM_3IMERGHHL.07') # this is the default version

# [1] 13
# This values agrees with the Snow phase report

Note that the GPM IMERG algorithm has deprecated the version 6 as of late 2024, see here

---

Example workflow:

Get modeled meteorological variables, harnessing data from the HADS, LCD, and WCC met networks.

# Define the static vars
met_networks = "ALL" # Call the HADS, LCD, and WCC stations
degree_filter = 1 # 1º radius

meteo <- rainOrSnowTools::access_meteo(networks         = met_networks,
                                       datetime_utc_obs = datetime,
                                       lon_obs          = lon,
                                       lat_obs          = lat,
                                       deg_filter       = degree_filter)

# Output: 387 observations of 7 variables
# head(meteo)
#      id            datetime temp_air rh temp_dew temp_wet ppt
# 1 AENC2 2025-04-10 11:56:00 1.111111 57       NA       NA  NA
# 2 AENC2 2025-04-10 12:56:00 2.222222 57       NA       NA  NA
# 3 BAWC2 2025-04-10 12:22:00 2.777778 33       NA       NA  NA
# 4 BAWC2 2025-04-10 13:22:00 4.444444 31       NA       NA  NA
# 5 BCFC2 2025-04-10 11:42:00 1.666667 45       NA       NA  NA
# 6 BCFC2 2025-04-10 12:42:00 1.111111 47       NA       NA  NA
 
# QC the meteo data
meteo_qc <- rainOrSnowTools::qc_meteo(meteo)
## All data pass QC

# Subset the data to select data points closest in time, and average the measured value
meteo_subset <- rainOrSnowTools:::select_meteo(meteo_qc, datetime)

# Get unique station IDs from the "meteo_qc" dataframe
stations_to_gather <- unique(meteo_qc$id)

# Get metadata for each station ID
metadata <- rainOrSnowTools::gather_meta(stations_to_gather)

# `metadata` can get the number of data points from each network

#>      hads_counts lcd_counts wcc_counts
#>              45         10         27

# `meteo_subset` can get the number of data points for each met variable

#>     ppt    rh temp_air temp_dew temp_wet
#>       4    40       81       12       10

# Finally, get modeled meteorological variables
met_vars <- rainOrSnowTools::model_meteo(id           = "example",
                                         lon_obs      = lon,
                                         lat_obs      = lat,
                                         elevation    = elev,
                                         datetime_utc = datetime,
                                         meteo_df     = meteo_subset,
                                         meta_df      = metadata)

# Output of 37 variables, does not include the QC flags

# Temperature units = ºC and RH unit = %
dplyr::tibble(met_vars)
#> # A tibble: 1 × 37
#>   id      temp_air_idw_lapse_const temp_air_idw_lapse_var temp_air_nearest_sit…¹
#>   <chr>                      <dbl>                  <dbl>                  <dbl>
#> 1 example                     1.25                   1.02                 -0.686
#> # ℹ abbreviated name: ¹​temp_air_nearest_site_const
#> # ℹ 33 more variables: temp_air_nearest_site_var <dbl>, temp_air_avg_obs <dbl>,
#> #   temp_air_min_obs <dbl>, temp_air_max_obs <dbl>, temp_air_lapse_var <dbl>,
#> #   temp_air_lapse_var_r2 <dbl>, temp_air_lapse_var_pval <dbl>,
#> #   temp_air_n_stations <int>, temp_air_avg_time_gap <dbl>,
#> #   temp_air_avg_dist <dbl>, temp_air_nearest_id <chr>,
#> #   temp_air_nearest_elev <dbl>, temp_air_nearest_dist <dbl>, …