Nota precautoria sobre el uso de bases de datos de precipitación mensual en el sudoeste de la Patagonia
DOI:
https://doi.org/10.24215/16699513e178Palabras clave:
sur de Sudamérica, productos reticulados, CRU, ERA5-Land, word clim, andes australesResumen
La precipitación es una variable climática fundamental para el funcionamiento de los ecosistemas, ya que regula la disponibilidad hídrica y condiciona procesos ecológicos y productivos. En la región de Patagonia, la baja densidad de estaciones meteorológicas limita los estudios basados en datos observacionales, por lo que se recurre a bases de datos reticuladas regionales y globales como CRU, WorldClim, GPCC, y ERA5-Land, entre otras. Sin embargo, la calidad de estas fuentes puede verse comprometida en áreas con orografía compleja y escasez de datos in situ. Este trabajo identifica un error significativo en los datos de precipitación mensual de las bases CRU y WorldClim para el sur de Chile y el oeste de la provincia de Santa Cruz (Argentina). A partir de agosto de 2009, en algunos puntos de la retícula los valores mensuales de precipitación no representan la variabilidad interanual, replicando de forma constante un ciclo medio anual. Este comportamiento se ve reflejado en el cambio porcentual del desvío estándar (SD) posterior a 2009, observándose caídas del mismo de hasta un 75% en la zona afectada. Por lo tanto, no se recomienda el uso de estas bases de datos de precipitación en la región afectada para periodos posteriores a agosto del 2009. En contraste, GPCC y ERA5-Land no presentan esta problemática, aunque ERA5-Land requiere una corrección de bias debido a un error sistemático en los datos. Estos resultados destacan la importancia de validar las fuentes de datos previamente a su aplicación en regiones con baja densidad de observaciones.
1. Introduction
Precipitation is a key climatic variable that regulates ecosystem functioning by determining soil water availability, directly influencing both ecological and productive processes (Noy-Meir, 1973). Precipitation dynamics affect various ecosystem factors such as plant growth regimes, insect life cycles, food availability for herbivores, predator activity, grassland primary productivity, and nutrient recycling (Pyle et al., 2021; Padda y Stahlschmidt, 2022; Wu et al., 2022). Moreover, precipitation is critical for assessing ecosystem responses to climate change and managing the risk of extreme events like droughts, floods, and wildfires (Almonacid et al., 2025; Johnston et al., 2020; Perri et al., 2024).
The Patagonian region is characterized by a very low density of meteorological stations, which hinders and limits research based on climate data (Skansi et al., 2013; Almonacid et al., 2021; Hurtado et al., 2024). To overcome this limitation, gridded precipitation products are widely used, including CRU, WorldClim, GPCC, and ERA5-based products. Among the most commonly used global precipitation datasets (WorldClim, CRU, ERA5, GPCC), a SCOPUS search using the keywords “precipitation” and “Patagonia” (conducted on April 23, 2025) found that many studies have used the first two: WorldClim (249 papers) and CRU (223), followed by ERA5 (215) and GPCC (60). Note that ERA5, being a reanalysis product of atmospheric variables at different heights, is not necessarily used as a precipitation proxy in all these studies. It is also important to note that WorldClim and CRU are interrelated datasets, as WorldClim is a statistical downscaling of CRU data (Fick y Hijmans, 2017).
These datasets are widely used in ecological, hydrological, and climate impact studies, often as the primary or only source of precipitation information in the region ( e.g,.(Morales et al., 2022; Ruiz Barlett et al., 2019; Fierke et al., 2024; Ricetti et al., 2024)(e.g., among others). However, their reliability is not uniform across space and time, particularly in regions with complex terrain and limited station density, where interpolation and reconstruction methods become increasingly uncertain (Kajornrit et al., 2011; Hurtado et al., 2021).
In this context, this study addresses a critical but previously undocumented issue in widely used gridded precipitation datasets. Specifically, we identify a structural artifact in CRU monthly precipitation data from August 2009 to the present, affecting southwestern Patagonia, which is subsequently propagated into WorldClim due to its dependence on CRU as input. This issue undermines the reliability of two of the most widely used climate datasets in a data-sparse region, with potential consequences for climate diagnostics and ecosystem impact studies. By documenting and characterizing this artifact, we provide essential guidance for the appropriate use and interpretation of gridded precipitation products.
2. Data and Methods
Four precipitation datasets were used in this study:
- Gridded precipitation data from WorldClim V2.1 with 1 km² spatial resolution (Fick y Hijmans, 2017).
- CRU TS 4.09 climate dataset with 0.5° spatial resolution (Harris et al., 2020).
- ERA5-Land precipitation data with 0.1° spatial resolution (Muñoz Sabater, 2019; Muñoz Sabater, 2019).
- GPCC Full Data Monthly Version 2022 with 0.25° spatial resolution (Schneider et al., 2022).
Both CRU and GPCC are based on interpolated observations. ERA5-Land is a reanalysis product based on ERA5 but with improved spatial resolution.The analyzed spatial domain is between 76°W-66°W and 45°S-55°S, encompassing most of Patagonia. For illustrative time series analysis (Fig. 1), we selected a representative grid cell located at 50°S, 75°W, chosen due to its central location within the region where the artifact is most pronounced, and considered the time period from 2000 to 2020 (21 years).
To assess the spatial cover of the data artifact in the CRU dataset, the standard deviation (SD) change was evaluated. The proportional SD change before and after August 2009 was computed as:
SD%Dif = (SD[period2]-SD_[period1])/SD_[period1] (Eq. 1)
Where period 1 spans from Jan 1994 to July 2009, and period 2 spans from August 2009 to December 2024. These two periods of equivalent length were selected to balance the amount of months (data length) after and before the detected issue (1 August 2009).
3. Results
Figure 1 shows monthly precipitation time series for southern Chile (grid point 50°S–75°W) from the four datasets. A clear change in precipitation behavior is observed in CRU and WorldClim datasets, starting in August 2009 (red dashed line). From that date onward, monthly precipitation values are constant for each calendar month (e.g., all Januaries have the same value, all Februaries as well, etc.). The time series exhibits inherent fluctuations until August 2009, after which it shifts to repeating its mean annual cycle, no longer capturing real-world variations. ERA5-Land and GPCC do not show this behavior. Moreover, as WorldClim is based on CRU data, it propagated the same error. Therefore, further analysis focuses solely on CRU.
To assess the spatial extent of the problem, the percentage change in standard deviation (SD) was calculated (Fig. 2). As the mean annual cycle is repeated, natural variability is lost, resulting in a significant drop in variance. A large region shows SD reductions between 60–75% near the grid point shown in Figure 1. Generally, throughout Patagonia, changes are under 15% (white on the map), with reductions greater than 30% only between 47.5°S and 52.5°S and west of 71.5°W. ERA5-Land and GPCC do not show changes greater than 30% and 45% respectively, and always under 15% in Chile and Argentinean Andes for ERA5-Land (not shown).
This decline in variability, with repetition of the annual cycle in certain grid points, may be due to a sharp drop in the number of rain gauge stations used by CRU (Fig. 3). Until 2000, around 8 weather stations were used throughout most of the region, and between 4 and 6 stations in the fjords and Tierra del Fuego. By 2010, the number of stations dropped drastically, reaching zero in the region where the problem was identified. In contrast, GPCC did not show a significant change in the number of stations used (not shown).
4. Discusion and Conclusions
This study identifies a previously undocumented structural error in CRU monthly precipitation data that is propagated into WorldClim due to its reliance on CRU as input. The issue affects a geographically constrained but highly sensitive region of southwestern Patagonia (47.5°S–52.5°S, west of 71.5°W), encompassing the Northern and Southern Ice Fields—key components of the regional hydrological system and major freshwater reserves with strong ecological and societal relevance (Carrasco-Escaff et al., 2023).
The presence of this artifact has important scientific implications. It undermines the reliability of two of the most widely used global precipitation datasets in a region where in situ observations are sparse and where gridded products are frequently used to support climate diagnostics, hydrological assessments, and ecosystem studies. As a result, analyses based on these datasets after August 2009 may be systematically biased, potentially affecting interpretations of precipitation variability, trends, and extremes. This is particularly critical in southwestern Patagonia, where precipitation strongly controls glacier mass balance, river discharge, and terrestrial ecosystem dynamics.From an environmental and societal perspective, errors in precipitation datasets can propagate through impact assessment frameworks, affecting hydrological modelling, ecosystem analyses, and decision-making processes. The choice of reference climate data has been shown to introduce uncertainty comparable to, or even exceeding, that of climate models themselves, particularly in data-sparse regions (Tarek et al., 2021). These uncertainties directly affect water resource assessments, as inaccuracies in hydroclimatic inputs propagate through modelling chains and can lead to substantial misestimation of system performance and risk, including errors in the evaluation of droughts and water availability (McInerney et al., 2024). In parallel, precipitation uncertainty can strongly influence ecological applications, as it underpins biome classification, biodiversity assessments, and the evaluation of climate change impacts on ecosystems (Navarro et al., 2024). Furthermore, differences among precipitation datasets can significantly alter estimates of hydrological ecosystem services and their beneficiaries, with direct implications for water management, hydroelectric planning, and policy-relevant decisions (Van Soesbergen y Mulligan, 2018). Therefore, reliable baseline precipitation information is essential to ensure robust environmental assessments and informed adaptation strategies under climate variability and change.
Based on these findings, the use of CRU and WorldClim datasets is not recommended for this region for periods after August 2009. Furthermore, the marked reduction in station density in southern Patagonia introduces additional uncertainty, suggesting that even outside the identified artifact, temporal variability may be artificially distorted.Among the alternative datasets analysed, GPCC and ERA5-Land do not exhibit this artifact and therefore represent preferable options. However, ERA5-Land presents a known precipitation bias in the region (Hurtado et al., 2024; Perri et al., 2024), requiring the application of bias correction methods. ERA5 has shown good performance in southern Patagonia (Almonacid et al., 2021), while GPCC—although not yet evaluated in this specific region—has shown weaker performance in northern Patagonia (Hurtado et al., 2024). This highlights the need for a dedicated evaluation of GPCC in southern Patagonia before its widespread application.Finally, these results underscore the importance of continuous validation of widely used gridded climate datasets, particularly in data-sparse and environmentally critical regions. Future versions of CRU and WorldClim should be carefully assessed to determine whether this issue has been resolved, ensuring the reliability of these products for scientific and applied purposes.
Authors´ contributions
Referencias
Descargas
Referencias
Almonacid, L., Pessacg, N. L., Diaz, B. G., Bonfili, O. and Peri, P. L. (2021). Nueva base de datos reticulada de precipitación para la provincia de Santa Cruz, Argentina. Meteorologica, 47(1), e010-e010. https://doi.org/10.24215/1850468xe010
Almonacid, L., Pessacg, N., Diaz, B. G. and Peri, P. L. (2025). Caracterización espacio-temporal de las sequías meteorológicas en Santa Cruz en el contexto de cambio climático. Meteorológica, 50, 035. https://doi.org/10.24215/1850468xe035
Carrasco-Escaff, T., Rojas, M., Garreaud, R. D., Bozkurt, D. and Schaefer, M. (2023). Climatic control of the surface mass balance of the Patagonian Icefields. The Cryosphere, 17(3), 1127-1149. https://doi.org/10.5194/tc-17-1127-2023
Fick, S. E. and Hijmans, R. J. (2017). WorldClim 2: New 1‐km spatial resolution climate surfaces for global land areas. International Journal of Climatology, 37(12), 4302-4315. https://doi.org/10.1002/joc.5086
Fierke, J., Joelson, N. Z., Loguercio, G. A., Putzenlechner, B., Simon, A., Wyss, D., Kappas, M. and Walentowski, H. (2024). Assessing uncertainty in bioclimatic modelling: A comparison of two high-resolution climate datasets in northern Patagonia. Regional Environmental Change, 24(3). https://doi.org/10.1007/s10113-024-02278-5
Harris, I., Osborn, T. J., Jones, P. and Lister, D. (2020). Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Scientific Data, 7(1). https://doi.org/10.1038/s41597-020-0453-3
Hurtado, S. I., Perri, D. V., Calianno, M., Martin-Albarracin, V. L. and Easdale, M. H. (2024). Monthly gridded precipitation databases performance evaluation in North Patagonia, Argentina. Theoretical and Applied Climatology, 155(9), 8771-8783. https://doi.org/10.1007/s00704-024-05153-9
Hurtado, S. I., Zaninelli, P. G., Agosta, E. A. and Ricetti, L. (2021). Infilling methods for monthly precipitation records with poor station network density in Subtropical Argentina. Atmospheric Research, 254, 105482. https://doi.org/10.1016/j.atmosres.2021.105482
Johnston, L. M., Wang, X., Erni, S., Taylor, S. W., McFayden, C. B., Oliver, J. A., Stockdale, C., Christianson, A., Boulanger, Y., Gauthier, S., Arseneault, D., Wotton, B. M., Parisien, M.-A. and Flannigan, M. D. (2020). Wildland fire risk research in Canada. Environmental Reviews, 28(2), 164-186. https://doi.org/10.1139/er-2019-0046
Kajornrit, J., Wong, K. W. and Fung, C. C. (2011). Estimation of missing rainfall data in northeast region of Thailand using spatial interpolation methods. Australian Journal of Intelligent Information Processing Systems, 13(1), 1-12.
McInerney, D., Thyer, M., Kavetski, D., Westra, S., Maier, H. R., Shanafield, M., Croke, B., Gupta, H., Bennett, B. and Leonard, M. (2024). Neglecting hydrological errors can severely impact predictions of water resource system performance. Journal of Hydrology, 634, 130853. https://doi.org/10.1016/j.jhydrol.2024.130853
Morales, C. L., Montalva, J., Arbetman, M. P., Aizen, M. A., Martins, A. C. and Silva, D. P. (2022). Does climate change influence the current and future projected distribution of an endangered species? The case of the southernmost bumblebee in the world. Journal of Insect Conservation, 26(2), 257-269. https://doi.org/10.1007/s10841-022-00384-5
Muñoz Sabater, J. (2019). ERA5-Land monthly averaged data from 1981 to present [Dataset]. In Climate Data Store. https://doi.org/10.24381/CDS.68D2BB30
Muñoz Sabater, J. (2021). ERA5-Land hourly data from 1950 to 1980 [Dataset]. In Climate Data Store (CDS). https://doi.org/10.24381/CDS.E2161BAC
Navarro, A., Lee, G., Martín, R. and Tapiador, F. J. (2024). Uncertainties in measuring precipitation hinders precise evaluation of loss of diversity in biomes and ecotones. Climate and Atmospheric Science, 7(1). https://doi.org/10.1038/s41612-024-00581-w
Noy-Meir, I. (1973). Desert ecosystems: Environments and producers. Annual Review of Ecology and Systematics, 4, 25-51. https://www.jstor.org/stable/2096803
Padda, S. S. and Stahlschmidt, Z. R. (2022). Evaluating the effects of water and food limitation on the life history of an insect using a multiple-stressor framework. Oecologia, 198(2), 519-530. https://doi.org/10.1007/s00442-022-05115-w
Perri, D. V., Hurtado, S. I., Bruzzone, O. and Easdale, M. H. (2024). Optimal automatic enhanced ERA5 daily precipitation data for environmental and agricultural monitoring tools in scarce data regions. Theoretical and Applied Climatology, 155(3), 1847-1856. https://doi.org/10.1007/s00704-023-04730-8
Pyle, L. A., Sheley, R. L. and James, J. J. (2021). Timing and duration of precipitation pulses and interpulses influence seedling recruitment in the Great Basin. Rangeland Ecology & Management, 75, 112-118. https://doi.org/10.1016/j.rama.2020.12.004
Ricetti, L., Hurtado, S. I., Agosta Scarel, E. and Cesanelli, A. (2024). Variabilidad del caudal del río Neuquén en las fases de su ciclo anual y su relación con índices climáticos. Meteorologica, 49, e026. https://doi.org/10.24215/1850468xe026
Ruiz Barlett, T., Martin, G. M., Laguna, M. F., Abramson, G. and Monjeau, A. (2019). Climatic constraints and the distribution of Patagonian mice. Journal of Mammalogy, 100(6), 1979-1991. https://doi.org/10.1093/jmammal/gyz149
Schneider, U., Hänsel, S., Finger, P., Rustemeier, E. and Ziese, M. (2022). GPCC Full Data Monthly Version 2022 at 0.25°: Monthly Land-Surface Precipitation from Rain-Gauges built on GTS-based and Historic Data [Dataset]. In Deutscher Wetterdienst. https://doi.org/10.5676/DWD_GPCC/FD_M_V2022_025
Skansi, M. de los M., Brunet, M., Sigró, J., Aguilar, E., Arevalo Groening, J. A., Bentancur, O. J., Castellón Geier, Y. R., Correa Amaya, R. L., Jácome, H., Malheiros Ramos, A., Oria Rojas, C., Pasten, A. M., Sallons Mitro, S., Villaroel Jiménez, C., Martínez, R., Alexander, L. V. and Jones, P. (2013). Warming and wetting signals emerging from analysis of changes in climate extreme indices over South America. Global Planet Change, 100, 295-307. https://doi.org/10.1016/j.gloplacha.2012.11.004
Tarek, M., Brissette, F. and Arsenault, R. (2021). Uncertainty of gridded precipitation and temperature reference datasets in climate change impact studies. Hydrology and Earth System Sciences, 25(6), 3331-3350. https://doi.org/10.5194/hess-25-3331-2021
Van Soesbergen, A. and Mulligan, M. (2018). Uncertainty in data for hydrological ecosystem services modelling: Potential implications for estimating services and beneficiaries for the CAZ Madagascar. Ecosystem Services, 33, 175-186. https://doi.org/10.1016/j.ecoser.2018.08.005
Wu, Q., Zeng, J. and Wu, K. (2022). Research and application of crop pest monitoring and early warning technology in China. Frontiers of Agricultural Science and Engineering, 9(1), 19-36. https://doi.org/10.15302/j-fase-2021411
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2026 Santiago Hurtado, Santino Adduca, Natalia Pessacg, Leandro Baltasar Díaz

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial-CompartirIgual 4.0.
Acorde a estos términos, el material se puede compartir (copiar y redistribuir en cualquier medio o formato) y adaptar (remezclar, transformar y crear a partir del material otra obra), siempre que a) se cite la autoría y la fuente original de su publicación (revista y URL de la obra), b) no se use para fines comerciales y c) se mantengan los mismos términos de la licencia.
La revista publica los artículos bajo esta licencia desde 2019 (Vol. 118 número 2).
Previo a esta fecha, los artículos se publicaron bajo una licencia Creative Commons Atribución (CC BY)


























