Access Canadian Historical Climate Data from R. The Government of Canada's Historical Climate Data website provides access to hourly, daily, and monthly weather records for stations throughout Canada.
These are raw data that have undergone some quality control, but issues such as changes in station location are unmanged; the data for the original stationID stops at a certain point and a new stationID continues recording. For a more curated data set for climate change research at broad spatial and temporal scales see the Adjusted and Homogenized Canadian Climate Data (AHCCD).
canadaHCD is still under active development towards a 0.1 release. In the meantime, if you wish to use the package, please install it from this github repo, which is most easily achieved using Hadley Wickham's devtools package:
## install.packages("devtools")
devtools::install_github("gavinsimpson/canadaHCD")Say I'm interested in climate data for stations in Yellowknife, I can search for all known stationIDs with "Yellowknife" in their name using find_station()
library("canadaHCD")
find_station("Yellowknife")
#> # A tibble: 7 x 5
#> Name Province StationID LatitudeDD
#> <fct> <fct> <fct> <dbl>
#> 1 YELLOWKNIFE A Northwest Territories 1706 62.5
#> 2 YELLOWKNIFE A Northwest Territories 51058 62.5
#> 3 YELLOWKNIFE OVERLAP Northwest Territories 53898 62.5
#> 4 YELLOWKNIFE-HENDERSON Northwest Territories 45467 62.4
#> 5 YELLOWKNIFE CON MINE Northwest Territories 8949 62.5
#> 6 YELLOWKNIFE CS Northwest Territories 27338 62.5
#> 7 YELLOWKNIFE HYDRO Northwest Territories 1707 62.7
#> LongitudeDD
#> <dbl>
#> 1 -114.
#> 2 -114.
#> 3 -114.
#> 4 -114.
#> 5 -114.
#> 6 -114.
#> 7 -114.To download the monthly HCD from YELLOWKNIFE HYDRO I can use hcd_monthly(), providing it with the StationID for that particular weather station
yh <- hcd_monthly(1707)The data are returned as a tibble (a tbl_df), which shows a compact version of the data frame.
yh
#> # A tibble: 690 x 16
#> Station StationID Longitude Latitude Date MaxTemp MinTemp MeanTemp
#> <chr> <chr> <dbl> <dbl> <yea> <dbl> <dbl> <dbl>
#> 1 YELLOW… 2204200 -114. 62.7 Jan … NA NA NA
#> 2 YELLOW… 2204200 -114. 62.7 Feb … -16.2 -26.4 -21.3
#> 3 YELLOW… 2204200 -114. 62.7 Mar … -14.4 -29.4 -21.9
#> 4 YELLOW… 2204200 -114. 62.7 Apr … 1.2 -12.3 -5.6
#> 5 YELLOW… 2204200 -114. 62.7 May … 9.3 -3.6 2.9
#> 6 YELLOW… 2204200 -114. 62.7 Jun … 17.6 2.4 10
#> 7 YELLOW… 2204200 -114. 62.7 Jul … 20.6 9.6 15.1
#> 8 YELLOW… 2204200 -114. 62.7 Aug … 18.9 7.2 13.1
#> 9 YELLOW… 2204200 -114. 62.7 Sep … 10.9 2 6.5
#> 10 YELLOW… 2204200 -114. 62.7 Oct … 6.1 -1.9 2.1
#> ExtremeHigh ExtremeLow TotalRain TotalSnow TotalPrecip LastSnowGrnd
#> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
#> 1 NA NA NA NA NA NA
#> 2 1.1 -44.4 0 9.4 9.4 NA
#> 3 -3.3 -40.6 0 2.8 2.8 NA
#> 4 12.2 -31.7 0 18 18 NA
#> 5 20 -11.7 9.9 2.8 12.7 NA
#> 6 27.2 -1.7 4.8 0 4.8 NA
#> 7 27.2 4.4 36.6 0 36.6 NA
#> 8 27.2 1.7 17.8 0 17.8 NA
#> 9 18.3 -6.1 5.8 2.8 8.6 NA
#> 10 17.2 -15.6 19.1 5.1 24.1 NA
#> # … with 680 more rows, and 2 more variables: MaxGustDir <int>,
#> # MaxGustSpeed <chr>You should be able to work with these objects mostly as if they were data frames.
Allthough not yet exposed through any functions in the package, you can access a snapshot of the station metadata via the canadaHCD:::station_data data frame.
canadaHCD:::station_data
#> # A tibble: 12,565 x 19
#> Name Province ClimateID StationID WMOID TCID LatitudeDD LongitudeDD
#> <fct> <fct> <fct> <fct> <fct> <fct> <dbl> <dbl>
#> 1 ACTI… British… 1010066 14 <NA> <NA> 48.9 -123.
#> 2 ALBE… British… 1010235 15 <NA> <NA> 48.4 -123.
#> 3 BAMB… British… 1010595 16 <NA> <NA> 48.6 -124.
#> 4 BEAR… British… 1010720 17 <NA> <NA> 48.5 -124
#> 5 BEAV… British… 1010774 18 <NA> <NA> 48.5 -123.
#> 6 BECH… British… 1010780 19 <NA> <NA> 48.3 -124.
#> 7 BREN… British… 1010960 20 <NA> <NA> 48.6 -123.
#> 8 BREN… British… 1010961 21 <NA> <NA> 48.6 -123.
#> 9 BREN… British… 1010965 22 <NA> <NA> 48.6 -123.
#> 10 BEAR… British… 1010PJR 9634 <NA> <NA> 48.5 -124.
#> Latitude Longitude Elevation FirstYear LastYear HourlyFirstYr
#> <int> <int> <dbl> <int> <int> <int>
#> 1 4.85e8 -1.23e9 4 1984 1996 NA
#> 2 4.82e8 -1.23e9 17 1971 1995 NA
#> 3 4.84e8 -1.23e9 85.3 1961 1980 NA
#> 4 4.83e8 -1.24e9 350. 1910 1971 NA
#> 5 4.83e8 -1.23e9 61 1894 1952 NA
#> 6 4.82e8 -1.23e9 12.2 1956 1966 NA
#> 7 4.84e8 -1.23e9 38 1987 1997 NA
#> 8 4.83e8 -1.23e9 30.5 1972 1980 NA
#> 9 4.83e8 -1.23e9 91.4 1960 1970 NA
#> 10 4.83e8 -1.24e9 419 NA NA NA
#> # … with 12,555 more rows, and 5 more variables: HourlyLastYr <int>,
#> # DailyFirstYr <int>, DailyLastYr <int>, MonthlyFirstYr <int>,
#> # MonthlyLastYr <int>If we wanted to know which resolutions of data were available for the YELLOWKNIFE HYDRO station, we can extract certain columns from the station data object
id <- grep("YELLOWKNIFE HYDRO", canadaHCD:::station_data$Name)
vars <- c("HourlyFirstYr", "HourlyLastYr", "DailyFirstYr", "DailyLastYr", "MonthlyFirstYr",
"MonthlyLastYr")
canadaHCD:::station_data[id, vars]
#> # A tibble: 1 x 6
#> HourlyFirstYr HourlyLastYr DailyFirstYr DailyLastYr MonthlyFirstYr
#> <int> <int> <int> <int> <int>
#> 1 NA NA 1943 2000 1943
#> MonthlyLastYr
#> <int>
#> 1 2000The output shows that this station has no hourly data, but daily and monthly data sets exist.