tidyhydat: An Introduction"

knitr::opts_chunk$set(echo = TRUE, 
                      warning = FALSE, 
                      message = FALSE, 
                      eval = nzchar(Sys.getenv("hydat_eval")),
                      fig.width=7, fig.height=7)

Package loading

In addition to tidyhydat, this vignette makes use of the dplyr package for data manipulations and ggplot2 for plotting.


tidyhydat package

This vignette will outline a few key options that will hopefully make tidyhydat useful.

HYDAT download

To use many of the functions in the tidyhydat package you will need to download a version of the HYDAT database, Environment and Climate Change Canada's database of historical hydrometric data then tell R where to find the database. Conveniently tidyhydat does all this for you via:


This downloads the most recent version of HYDAT and then saves it in a location on your computer where tidyhydat's function will look for it. Do be patient though as this takes a long time! To see where HYDAT was saved you can run hy_dir(). Now that you have HYDAT downloaded and ready to go, you are all set to begin some hydrologic analysis.


Most functions in tidyhydat follow a common argument structure. We will use the hy_daily_flows() function for the following examples though the same approach applies to most functions in the package (See ls("package:tidyhydat") for a list of exported objects). Much of the functionality of tidyhydat originates with the choice of hydrometric stations that you are interested in. A user will often find themselves creating vectors of station numbers. There are several ways to do this.

The simplest case is if you would like to extract only station. You can supply this directly to the station_number argument:

hy_daily_flows(station_number = "08LA001")

Another method is to use hy_stations() to generate your vector which is then given the station_number argument. For example, we could take a subset for only those active stations within Prince Edward Island (Province code:PE) and then create vector for hy_daily_flows():

PEI_stns <- hy_stations() %>%
  filter(HYD_STATUS == "ACTIVE") %>%
  filter(PROV_TERR_STATE_LOC == "PE") %>%


hy_daily_flows(station_number = PEI_stns)

We can also merge our station choice and data extraction into one unified pipe which accomplishes a single goal. For example if for some reason we wanted all the stations in Canada that had the name "Canada" in them we unify that selection and data extraction process into a single pipe:

search_stn_name("canada") %>%
  pull_station_number() %>%

We saw above that if we were only interested in a subset of dates we could use the start_date and end_date arguments. A date must be supplied to both these arguments in the form of YYYY-MM-DD. If you were interested in all daily flow data from station number "08LA001" for 1981, you would specify all days in 1981 :

hy_daily_flows(station_number = "08LA001", 
               start_date = "1981-01-01", end_date = "1981-12-31")

This generally outlines the usage of the HYDAT functions within tidyhydat.

Real-time functions

In addition to the approved and vetted data in the HYDAT database ECCC also offers unapproved data that is subject to revision. tidyhydat provides three functions to access these data sources. Remember these are unapproved data and should treated as such:

Not every stations is currently part of the real-time network. Therefore realtime_stations() points to a (hopefully) updated ECCC data file of active real-time stations. We can use the realtime_stations() functionality to get a vector of stations by jurisdiction. For example, we can choose all the stations in Prince Edward Island using the following:

realtime_stations(prov_terr_state_loc = "PE")

hy_stations() and realtime_stations() perform similar tasks albeit on different data sources. hy_stations() extracts directly from HYDAT. In addition to real-time stations, hy_stations() outputs discontinued and non-real-time stations:

hy_stations(prov_terr_state_loc = "PE")

This is contrast to realtime_stations() which downloads all real-time stations. Though this is not always the case, it is best to use realtime_stations() when dealing with real-time data and hy_stations() when interacting with HYDAT. It is also appropriate to filter the output of hy_stations() by the REAL_TIME column.

Meterological Service of Canada datamart - realtime_dd()

To download real-time data using the datamart we can use approximately the same conventions discussed above. Using realtime_dd() we can easily select specific stations by supplying a station of interest:

realtime_dd(station_number = "08LG006")

Another option is to provide simply the province as an argument and download all stations from that province:

realtime_dd(prov_terr_state_loc = "PE")

Search functions

You can also make use of auxiliary functions in tidyhydat called search_stn_name() and search_stn_number() to look for matches when you know part of a name of a station. For example:


Similarly, search_stn_number() can be useful if you are interested in all stations from the 08MF sub-sub-drainage:


Using joins

Sometimes it is required to make use of information from two tables from HYDAT. In some cases, we need to combine the information into one table using a common column. Here we will illustrate calculating runoff by combining the hy_stations tables with the hy_daily_flows table by the STATION_NUMBER column:

stns <- c("08NH130", "08NH005")
runoff_data <- hy_daily_flows(station_number = stns, start_date = "2000-01-01") %>%
    hy_stations(station_number = stns) %>%
    by = "STATION_NUMBER") %>%
  ## conversion to mm/d
  mutate(runoff = Value / DRAINAGE_AREA_GROSS * 86400 / 1e6 * 1e3) 

ggplot(runoff_data) + 
  geom_line(aes(x = Date, y = runoff, colour = STATION_NAME)) +
  labs(y = "Mean daily runoff [mm/d]") +


Copyright 2017 Province of British Columbia

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at


Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
See the License for the specific language governing permissions and
limitations under the License.

Try the tidyhydat package in your browser

Any scripts or data that you put into this service are public.

tidyhydat documentation built on May 18, 2021, 1:09 a.m.