Canadian Homogenized Precipitation – Version 2 (CanHomP V2)
Published by Environment and Climate Change Canada | Environnement et Changement climatique Canada
About this dataset
Canadian Homogenized Precipitation – Version 2 (CanHomP V2) The CanHomP V2 datasets were developed for climate trend analysis and include long-term monthly and daily precipitation series for 425 stations across Canada (Wang et al., 2026; Wang & Feng, 2026). Key improvements in Version 2 over its predecessor, CanHomPmlyV1 (Wang et al., 2023), include: • Expanded and improved source datasets, incorporating adjusted data from automated gauge stations and the Collaborative Rain, Hail and Snow (CoCoRaHS) network (https://www.cocorahs.org/canada.aspx) for recent decades. • Additional quality control procedures to remove false zeros in station data records for data-sparse regions or periods. • Enhanced changepoint detection, including tests on both untransformed and log-transformed data, and improved methods for identifying station-joining and variance inhomogeneities. • Use of more complete metadata for better accuracy. • Improved adjustment procedures to eliminate inhomogeneity by adjusting untransformed data series. • Development of homogenized daily precipitation series consistent with corresponding monthly totals (Wang & Feng, 2026). Key processing steps for both V1 and V2 include: • Merging observations from nearby sites to create long records, primarily using updated Adjusted Daily Rainfall and Snowfall data corrected for known issues such as unrealistic snow-water equivalent conversion, gauge wetting loss, and wind-induced undercatch (Wang et al., 2017). • Infilling data gaps using advanced spatial modeling of available data for the gap period. • Comprehensive quality control of data (Cheng et al., 2024). • Detecting non-climatic shifts using homogeneity tests with reference and station metadata. • Using up to four best neighbor stations as references, along with data derived from advanced spatial modeling of adjusted precipitation (MacDonald et al., 2021) and the Twentieth Century Reanalysis (20CRv3) ensemble-mean series of monthly precipitation (Slivinski et al., 2019). • Applying Quantile-Matching (QM) adjustments without reference data to correct non-climatic changes (Wang et al., 2026; Wang et al., 2023; Wang et al., 2010). Wang et al. (2023) also developed a dynamic QM adjustment method that uses the most strongly correlated homogeneous segment of nearby station data as a reference. They found that when station density is too low to identify suitable reference stations, adjustments without a reference could perform better than those with one, although results are similar for most stations. This approach may not be applicable to precipitation datasets with much higher station density or to variables with greater spatial coherence (e.g., surface air temperature). The QM method adjusts the entire distribution of data in one segment to match another (Wang et al., 2026; Wang et al., 2023; Wang & Feng, 2013; Wang et al., 2010), ensuring that distribution changes—including variance shifts—at identified changepoints are homogenized. Differences from AHCCD adjusted precipitation data • CanHomP V2: Adjusted, gap-filled, homogenized monthly and daily precipitation data for 425 core stations for the period up to 2023. • AHCCD adjusted precipitation: Adjusted but unhomogenized monthly and daily precipitation data for 464 manual stations for the period up to 2012 (Mekis & Vincent, 2011). References Wang, X. L., Feng, Y., Zwiers, F. W., & Cheng, V. Y. S. (2026). Precipitation trends in version 2 of Canadian homogenized monthly precipitation dataset. Atmosphere-Ocean, 1-16. https://doi.org/10.1080/07055900.2026.2617861. Wang, X. L., & Feng, Y. (2026). Observed trends in precipitation extreme indices as inferred from a homogenized daily precipitation dataset for Canada. Weather and Climate Extremes, 51, 100860. https://authors.elsevier.com/sd/article/S2212-0947(26)00011-3. Wang, X.L, Y. Feng, V. Y. S. Cheng, H. Xu, 2023: Observed precipitation trends inferred from Canada’s homogenized monthly precipitation dataset, J. Clim., 36, 7957-79