Abstract
Eddy covariance sites are ideally suited for the study of extreme events on ecosystems as they allow the exchange of trace gases and energy fluxes between ecosystems and the lower atmosphere to be directly measured on a continuous basis. However, standardized definitions of hydroclimatic extremes are needed to render studies of extreme events comparable across sites. This requires longer datasets than are available from on-site measurements in order to capture the full range of climatic variability. We present a dataset of drought indices based on precipitation (Standardized Precipitation Index, SPI), atmospheric water balance (Standardized Precipitation Evapotranspiration Index, SPEI), and soil moisture (Standardized Soil Moisture Index, SSMI) for 101 ecosystem sites from the Integrated Carbon Observation System (ICOS) with daily temporal resolution from 1950 to 2021. Additionally, we provide simulated soil moisture and evapotranspiration for each site from the Mesoscale Hydrological Model (mHM). These could be utilised for gap-filling or long-term research, among other applications. We validate our data set with measurements from ICOS and discuss potential research avenues.
Similar content being viewed by others
Background & Summary
Europe has been hit by a series of droughts and hot spells in recent years1, with climatic conditions in the last decade among the most extreme in more than 250 years2. Quantifying the impact of these events on the terrestrial carbon sequestration capacity is crucial for understanding how future climate change may alter the terrestrial carbon cycle3,4. This requires direct measurements of the exchange of greenhouse gases between the lower atmosphere and the terrestrial surface performed using the eddy covariance technique5. Due to the high technical investment and site requirements alongside the complexity of the measurements, the application of eddy covariance is only possible at certain locations. Research infrastructure networks such as the Integrated Carbon Observation System (ICOS-RI) aim to ensure that measurements are representative and comparable through standardization and quality control6,7. ICOS also advances research on relevant topics. In response to the 2018 extreme compound drought and heat event in Europe, for example, the Drought-2018 initiative published an aggregated dataset of affected stations8, which strongly facilitated the study of the extreme event impacts9,10,11,12,13,14,15,16,17,18,19,20,21,22,23.
Despite the efforts of research infrastructure networks such as ICOS, there is still a lack of standardized datasets for quantifying drought severity. This partly arises from the lack of a uniform definition of drought severity24,25, preventing holistic assessment of extreme droughts and their impact26. For example, a meta-analysis of 564 ecological drought studies27 found that conditions in ~50% of all studies were within the range of normal climatic variability and therefore could not be classified as drought, and at least 30% of studies confused aridity with drought. This highlights the urgent need for uniform and standardized definitions.
To support research into drought impact on ecosystems, we provide a dataset of daily standardized drought indices tailored to each ICOS ecosystem station. We calculate three indices: the Standardized Precipitation Index (SPI)28,29, which is based on precipitation anomalies only, the Standardized Precipitation Evapotranspiration Index (SPEI30), which is based on a simplified water balance, and a Standardized Soil Moisture Index (SSMI), which is based on modelled soil moisture anomalies31. While a variety of other indices can be found in the literature24,32,33,34, probabilistic indices such as these allow standardized comparisons as they express the current conditions with respect to the long-term climatology of the respective location. Such standardized indices enable comparison even across larger spatial areas or between climate zones33,35.
While the probabilistic approach is a robust and widely used method34, it comes with a challenge. Robust estimation of the climatic distribution of the target variable requires 30 years of data at least28, and preferably as much as 50 to 80 years36. The first eddy covariance stations were established only in the late 1990s, with many stations being much younger. Therefore, in order to estimate the climatology, external data sources are necessary. Here, we combine long-term observational data with simulations from the mesoscale Hydrological Model (mHM)37,38 to provide a comprehensive drought description. Both observational and model data are compared to ICOS measurements to ensure that the data provided accurately represents the meteorological conditions at each site.
In summary, we provide a consistent long-term hydrometeorological database of drought indices that allows the comparative analysis of e.g. ecosystem responses based on standardized drought severity across different ICOS ecosystem stations. We additionally provide simulated long-term data of soil moisture and evapotranspiration, and demonstrate exemplary use-cases. We hope that the database, with its daily temporal resolution and a wide range of aggregation periods ranging from a few days to two years, will facilitate future studies on the impact of droughts on European ecosystems and ensure comparability of research results.
Methods
Data source and workflow
Drought indices were calculated based on the European Climate Assessment & Dataset (ECA&D) E-OBS gridded dataset v.25.0e39 and soil moisture simulations realized with the mesoscale Hydrologic Model v5.11.1 (mHM)37,38. The workflow is presented in Fig. 1. For each ICOS ecosystem site, the respective meteorological forcing data were extracted from the corresponding E-OBS grid cell (~10 × 10 km2) and used to calculate indices directly (in the case of SPI and SPEI) or to run mHM to derive soil moisture simulations for SSMI. mHM was set up with the same resolution as E-OBS and the digital elevation model from USGS40, the soil map from SOILGRIDS41, the land cover from ESA42 and LAI climatology from NASA Global Inventory, Monitoring, and Modelling Studies (GIMMS)43. The underlying parameterization of the European-wide mHM setup was based on established settings, which have been evaluated in previous studies2,44,45.
SPI and SPEI were calculated using nonparametric kernel density estimation (KDE) with aggregation times ranging from 5 to 365 days (in steps of 5 days) and 370 to 720 days (in steps of 10 days). No aggregation was performed for the SSMI because soil moisture itself naturally reflects past meteorological conditions. The SSMI was calculated based on two different mHM soil layers, one from the upper soil layer (SSMI_top, ~30 cm), reflecting shorter meteorological fluctuations, and the other considering the entire soil layer up to 2 m soil depth (SSMI_full, ~200 cm).
Figure 2 illustrates the spatial coverage of our data set. Stations were selected using the ICOS station network overview (https://www.icos-cp.eu/observations/station-network) and filtering by THEME = “ecosystem”. The most recent observational data product for eddy covariance fluxes from ICOS, the Warm Winter 2020 dataset46 (WW20 hereafter), was used as ground truth for the technical validation. Stations included in WW20, but not in the ICOS station network, were also used (indicated by the dark blue on the map). Stations in Greenland, Israel, Congo and French Guiana were removed as E-OBS does not cover those locations. In total, we provide drought indices for 101 stations in Europe, of which 74 are validated against ICOS observational data.
Standardized drought indices
Since its introduction in the 1990s28,29, the SPI has been widely applied as a robust method to compare precipitation anomalies objectively across different climate zones47,48,49,50, and is recommended by the World Meteorological Organization as a key drought indicator51. The SPEI30 is an extension of the SPI which additionally accounts for changes in Potential Evapotranspiration (PET), and is therefore better suited for the detection of droughts which are induced by increased air temperature34,52,53. In contrast to the SPI and SPEI, the SSMI is not based on observational data directly, as sufficient historical records for soil moisture do not exist. Probabilistic soil moisture indices are usually constructed from model-based soil moisture estimates54,55,56,57. The soil moisture index based on mHM, which is used in this study to simulate long-term soil moisture time series at the respective sites, is used in the German Drought Monitor58 and the South Asian Drought Monitor59 and has been validated and utilized for multiple applications60,61.
Probability-based indices are traditionally derived by fitting the variable of interest to a suitable probability density function (pdf). The corresponding standard normal deviation can be obtained from the resulting cumulative density probability (cdf) using the equiprobability transformation. The choice of the most suitable pdf is non-trivial and has been a matter of debate since the development of probability indices29,62,63,64. Instead of assuming a parametric distribution function, one can determine the density distribution from the empirical data itself65,66,67,68. The advantages of this approach are that no assumptions need to be made about the functional form, thus avoiding possible bias due to an inappropriate parametric model. Therefore, in this study we apply non-parametric kernel density estimation (KDE) for estimating the drought indices for all variables: P, P-PET, and SM.
We calculated daily drought indices (i.e., SPI, SPEI, and SSMI) using KDE. For a given time series of the target variable \({X}_{i,j}^{k}\left(i=1,\ldots ,n\right)\) with j being the time step (here, daily) and k the number of previous time steps used for aggregation to the current time step, the kernel density for each j can be obtained using a Gaussian kernel by:
where n represents the sampling size and h the bandwidth. In kernel density estimation, the bandwidth selection is the highest source of uncertainty69. A common approach is unbiased cross-validation (CV)31,70:
where dropping xi when estimating f(x) (denoted by \({\widehat{f}}_{-i}\left(x| h\right)\)) yields the least-squares CV criterion. Finally, to facilitate the comparison with classic drought categories found in the literature (e.g. in Table 2), derived based on the standard scores, we transform the KDE-based quantiles into standardized normal scores. These standardized values can be obtained from the cumulative density function F(x) by an approximation of the normal quantiles71:
where
with P = 1-F(x) and C0 = 2.515517, C1 = 0.802853, C2 = 0.010328, d1 = 1.432788, d2 = 0.189269, d3 = 0.001308.
Kernel estimators are known to perform poorly for bounded random variables, e.g. precipitation, in which f(x) is only supported on R+ = (0,+∞), as kernels can extend past such boundaries, causing leakage of probability mass72. To avoid erroneous estimations for x = 0 precipitation, we extend the non-parametric estimation for the probability of zero precipitation, similarly to the original SPI method:
where q is the empirical probability of x = 0 precipitation.
Estimation of potential evapotranspiration
The potential evapotranspiration required for the SPEI determination was calculated using the Hargreaves-Samani equation (H-S)73:
where TG, TX and TN are the average, minimum and maximum temperature, respectively, and RG is the global radiation, i.e. the total of the shortwave solar radiation reaching the ground. Several methods have been proposed for estimating PET, ranging from simple empirical models with few parameters, such as the Thornthwaite model74, which requires only the daily or monthly average temperature, to more complex physical models, such as the Penman-Monteith (P-M) model75. The H-S model is a parsimonious option with low data demand and reasonable accuracy76, and was therefore chosen here. PET is also provided directly from E-OBS, based on the Makkink77 or the P-M model, but it is calculated based on an older E-OBS version and therefore covers a shorter time period than our dataset.
Global radiation is estimated based on latitude as proposed in the original SPEI publication30. Global radiation is also available from the E-OBS dataset, but with poorer spatial coverage and more gaps than the temperature variables. To have a consistent estimate of PET and soil moisture, we estimated global radiation as:
where dr represents the inverse relative distance between Earth and the sun, ω is the sunrise hour angle, lat is the latitude, δ is the sun declination in radians, and doy is the day of the year. The solar declination can be calculated as follows:
The inverse relative distance between Earth and the sun, corrected for the eccentricity of Earth’s orbit around the sun, can be calculated as follows:
and the sunrise hour angle as follows:
Data Records
The dataset can be obtained from a Zenodo repository78 and comprises four files for each of the 101 sites. “[site_name]_input.csv” contains the observational data extracted from E-OBS gridded dataset v.25.0e39, PET estimates as well as the simulated soil water storage and actual evapotranspiration from the mesoscale Hydrologic Model v5.11.1 (mHM)37,38; and the remaining files contain the drought indices (“SSMI_[site_name].csv”, “SPI_[site_name].csv”, “SPEI_[site_name].csv”). Details of the variables, their units and their origin are given in Table 1. For the SPI and SPEI, the file of each site contains the estimates for various aggregation times, ranging from 5 to 720 days in steps of 5 days from 5 to 365 and steps of 10 days from 370 to 720. Each data file has a daily temporal resolution and covers the time span from 1950 to 2021.
Technical Validation
The accuracy of drought indices depends largely on the reliability of the meteorological variables used for their calculation. The E-OBS data have been extensively used and validated39,79. Nevertheless, we compared the measurements at the ICOS stations with the extracted grid cells from E-OBS, with the results shown in Fig. 3a–d. The temperature variables show high agreement between E-OBS and ICOS data. Precipitation showed low error on average (in terms of normalized RMSE), but the calculated Pearson’s correlation coefficient is almost half of those obtained for the temperature variables. We find that smaller precipitation events show higher uncertainty. This is not unexpected, since precipitation is notoriously difficult to estimate at high spatial or temporal resolution80. However, since the indices are calculated using aggregated data from the last x days, this uncertainty is offset by the aggregation.
PET estimated with E-OBS data is generally higher than the observations from ICOS because the radiation is estimated from latitude only, and therefore does not include information on cloud cover. However, it has been shown that this is less relevant than the temporal consistency for SDI due to the standardization30,81, which is estimated by the R2 of 0.85. Figure 3e,f shows modelled water fluxes (namely, evapotranspiration (ET) and soil moisture) from mHM. Both modelled variables have low performance compared to the observations from ICOS. The mHM setup has a resolution of approximately 10 × 10 km2 and can therefore only approximately represent the conditions at specific locations. On the other hand, the soil moisture index does not require temporal aggregation, and its implications may be more straightforward because it is based directly on soil moisture information, while SPI and SPEI are only proxies for the water available to the ecosystem.
To demonstrate the information contained in the mHM output, we trained regression models for each ICOS site from the WW20 dataset using the eddy covariance fluxes from the sites as the target variable. More precisely, we trained regression models separately to predict net ecosystem exchange (NEE_VUT_MEAN), gross primary productivity (GPP_NT_VUT_MEAN), ecosystem respiration (RECO_NT_VUT_MEAN) and latent heat flux (LE_F_MDS) for the growing seasons (here May–September). In each model, we used shortwave incoming radiation (SW_IN_F) as a predictor and alternative soil moisture information as a covariate, namely soil water content measurements (from the ICOS sites, SWC_MDS_F_3) or simulated soil water content (from mHM, SM_top or SM_full). The regression models were trained using restricted cubic spline regression from the R-package “rms”82 with 5 knots per variable.
Figure 4 illustrates that the models achieve similar performance regardless of whether the soil moisture information provided comes from observations or simulation, and regardless of whether soil moisture is of higher or lower importance to the model. We assume that the comparable performance can be attributed to a similar level of uncertainty in the datasets, despite the fact that this uncertainty arises from different causes. Uncertainty in the results from mHM is may be due to the coarse resolution representing the soil moisture conditions at the specific sites less accurately, while the observed data from ICOS contains greater noise. For a fair comparison, we only used data from days in which soil moisture measurements were available. The great advantage of the simulations, however, is that they provide gap-free time series over many years.
Usage Notes
In the following, we illustrate how the derivation of drought indices from climate data facilitates their utility. An exemplary transformation of observational data to a drought index is shown in Fig. 5. Here, we calculate the SPEI with an aggregation time of 30 days (as an equivalent to the SPEI of one month) for July 1 for the ICOS station “FI-Hyy”. In other words, we aim to estimate the climatic conditions on July 1 based on meteorology from the previous 30 days. To do this, we sum the water balance for the last 30 days of each year for which we obtained data (here 1950–2021). This allows us to estimate the empirical distribution function (cf. Figure 5a) of the atmospheric water balance on July 1 with Eq. (1). Figure 5c shows how the histogram of the sample relates to the estimated probability function of the KDE from Eq. (1).
In the next step, the empirical probabilities are transformed equi-probabilistically into the standard normal probabilities using Eq. (3). For example, the sum of the water balance on July 1, 2004 was 0.78 mm in FI-Hyy. The corresponding cumulative probability is highlighted in Fig. 5a. After the transformation, the probability of a given summed water balance on July 1 can be read as the standard deviation of a normal distribution as shown in Fig. 5b. A cumulative water balance of 0.78 mm corresponds to a standard deviation of −0.796, and therefore does not classify as drought (cf. Table 2). Note that this applies only to July 1. The exact value of the water deficit is less important for the calculation of the SPEI than the probability of occurrence. Rain or water deficits of equal magnitude can correspond to drought or non-drought conditions depending on the time of year of occurrence. Therefore, these steps must be repeated for each day of the year to obtain a complete time series.
The results of the drought index allow the definition of drought categories. Different approaches can be found in the literature, which we summarise in Table 2. McKee28 originally proposed a division into mild (over −1), moderate (over −1.5), severe (over −2) and extreme (below −2) drought. Agnew83 criticized the fact that any sub-zero SDI (e.g. below average water availability) would be classified as a drought under this definition. They suggest a classification anchored on probabilities rather than deviations. In the Agnew’s classes, moderate, severe, and extreme droughts have probabilities of 20%, 10%, and 5%, respectively, while in the original classification, they have probabilities of 15.9%, 6.7%, and 2.3%, respectively (cf. Table 2). The two different classifications are visualized in Fig. 5, both for the probabilities and the standard deviation. The U.S. Drought Monitor84 uses similar probability classes to those Agnew proposed, with one additional class.
These classification differences, although seemingly minor, can have a strong impact on drought assessment. As an example, the daily time series of the proportion of ecosystem ICOS sites affected by drought is shown in Fig. 6. We used all three drought indices with an aggregation time of 365 days for the SPI and SPEI, and applied both classification schemes described above. During the recent 2018–2020 drought and heat wave in Europe, extreme drought affected up to 40–50% of all sites during the most intense period in 2019 according to the Agnew classification, compared to only 20–25% drought according to the McKee classification. Although the number of sites affected by drought is similar, the number of sites affected by “extreme” drought is very different. This emphasizes the importance of specifying the classification scheme used to characterise drought severity levels to facilitate the comparison of research results from different studies.
We anticipate that the daily temporal resolution of this dataset will enable more detailed examinations of ecosystem response to drought, as drought indices are predominantly used with monthly resolution in current research (cf. e.g. https://spei.csic.es/index.html). To get a first insight, we plotted the correlation between GPP and SPEI as a two-dimensional time series for several sites in Fig. 7. We applied a z-transformation of GPP for each day of the year to remove seasonal phenology effects, and then calculated a running Pearson’s correlation coefficient using a rolling window size of 14 days with all available aggregation levels of SPEI. Consequently, the x-axis represents the temporal development of GPP (as anomalies) during the growing season, while the y-axis represents the SPEI with increasing aggregation time, or in other words, an increasing amount of information on past conditions. As this is only an exemplary analysis of what could be done with the high-resolution dataset provided, rather than a comprehensive analysis, a more sophisticated analysis is needed to better understand how ecosystems respond to shorter and longer droughts.
In summary, we provide a long-term dataset with information on drought conditions for ICOS ecosystem sites in Europe, which additionally includes long-term simulations of soil moisture and evapotranspiration for each site. Depending on community feedback on the applicability of the data provided, we plan to update the dataset annually to provide a consistent dataset for drought research on European ecosystems. We believe that the use of the standardized drought indices provided in this dataset will improve the comparability of studies on the impacts of extreme events across the various ICOS ecosystem sites.
Code availability
Code for calculation of the drought indices as well as all statistical analysis within this publication is publicly available at the zenodo repository85. The mesoscale Hydrologic Model is an open source model and is available at https://mhm-ufz.org/.
References
Boergens, E., Güntner, A., Dobslaw, H. & Dahle, C. Quantifying the Central European Droughts in 2018 and 2019 With GRACE Follow-On. Geophysical Research Letters 47, e2020GL087285 (2020).
Rakovec, O. et al. The 2018– 2020 Multi-Year Drought Sets a New Benchmark in Europe. Earth’s Future 10, e2021EF002394 (2022).
Sippel, S. et al. Drought, Heat, and the Carbon Cycle: A Review. Current Climate Change Reports 4, 266–286 (2018).
Humphrey, V. et al. Soil moisture– atmosphere feedback dominates land carbon uptake variability. Nature 592, 65–69 (2021).
Foken, T., Aubinet, M. & Leuning, R. The Eddy Covariance Method. In Aubinet, M., Vesala, T. & Papale, D. (eds.) Eddy Covariance, 1–19 (Springer Netherlands, Dordrecht, 2012).
Franz, D. et al. Towards long-term standardised carbon and greenhouse gas observations for monitoring Europe’s terrestrial ecosystems: A review. International Agrophysics 32, 439–455 (2018).
Heiskanen, J. et al. The Integrated Carbon Observation System in Europe. Bulletin of the American Meteorological Society 103, E855–E872 (2022).
Drought 2018 Team and ICOS Ecosystem Thematic Centre. Drought-2018 ecosystem eddy covariance flux product for 52 stations in FLUXNET-Archive format. ICOS Carbon Portal https://doi.org/10.18160/YVR0-4898 (2020).
Bastos, A. et al. Impacts of extreme summers on European ecosystems: A comparative analysis of 2003, 2010 and 2018. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190507 (2020).
Beillouin, D., Schauberger, B., Bastos, A., Ciais, P. & Makowski, D. Impact of extreme weather conditions on European crop production in 2018. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190510 (2020).
El-Madany, T. S. et al. Drought and heatwave impacts on semi-arid ecosystems’ carbon fluxes along a precipitation gradient. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190519 (2020).
Fu, Z. et al. Sensitivity of gross primary productivity to climatic drivers during the summer drought of 2018 in Europe. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190747 (2020).
Gharun, M. et al. Physiological response of Swiss ecosystems to 2018 drought across plant types and elevation. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190521 (2020).
Gourlez de la Motte, L. et al. Non-stomatal processes reduce gross primary productivity in temperate forest ecosystems during severe edaphic drought. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190527 (2020).
Graf, A. et al. Altered energy partitioning across terrestrial ecosystems in the European drought year 2018. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190524 (2020).
Koebsch, F. et al. The impact of occasional drought periods on vegetation spread and greenhouse gas exchange in rewetted fens. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190685 (2020).
Kowalska, N. et al. Analysis of floodplain forest sensitivity to drought. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190518 (2020).
Peters, W., Bastos, A., Ciais, P. & Vermeulen, A. A historical, geographical and ecological perspective on the 2018 European summer drought. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190505 (2020).
Ramonet, M. et al. The fingerprint of the summer 2018 drought in Europe on ground-based atmospheric CO2 measurements. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190513 (2020).
Rinne, J. et al. Effect of the 2018 European drought on methane and carbon dioxide exchange of northern mire ecosystems. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190517 (2020).
Rödenbeck, C., Zaehle, S., Keeling, R. & Heimann, M. The European carbon cycle response to heat and drought as seen from atmospheric CO2 data for 1999– 2018. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190506 (2020).
Smith, N. E. et al. Spring enhancement and summer reduction in carbon uptake during the 2018 drought in northwestern Europe. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190509 (2020).
Thompson, R. L. et al. Changes in net ecosystem exchange over Europe during the 2018 drought based on atmospheric observations. Philosophical Transactions of the Royal Society B: Biological Sciences 375, 20190512 (2020).
Keyantash, J. & Dracup, J. A. The Quantification of Drought: An Evaluation of Drought Indices. Bulletin of the American Meteorological Society 83, 1167–1180 (2002).
Stahl, K. et al. The challenges of hydrological drought definition, quantification and communication: An interdisciplinary perspective. In Proceedings of IAHS, vol. 383, 291–295 (Copernicus GmbH, 2020).
McPhillips, L. E. et al. Defining Extreme Events: A Cross-Disciplinary Review. Earth’s Future 6, 441–455 (2018).
Slette, I. J. et al. How ecologists define drought, and why we should do better. Global Change Biology gcb.14747 (2019).
McKee, T. B., Doesken, N. J. & Kleist, J. The relationship of drought frequency and duration to time scales. Proceedings of the 8th Conference on Applied Climatology 17, 179–183 (1993).
Guttman, N. B. Accepting the Standardized Precipitation Index: A calculation algorithm. JAWRA Journal of the American Water Resources Association 35, 311–322 (1999).
Vicente-Serrano, S. M., Beguería, S. & López-Moreno, J. I. A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index. Journal of Climate 23, 1696–1718 (2010).
Samaniego, L., Kumar, R. & Zink, M. Implications of Parameter Uncertainty on Soil Moisture Drought Analysis in Germany. Journal of Hydrometeorology 14, 47–68 (2013).
Heim, R. R. A Review of Twentieth-Century Drought Indices Used in the United States. Bulletin of the American Meteorological Society 83, 1149–1166 (2002).
Zargar, A., Sadiq, R., Naser, B. & Khan, F. I. A review of drought indices. Environmental Reviews 19, 333–349 (2011).
Mukherjee, S., Mishra, A. & Trenberth, K. E. Climate Change and Drought: A Perspective on Drought Indices. Current Climate Change Reports 4, 145–163 (2018).
Mishra, A. K. & Singh, V. P. A review of drought concepts. Journal of Hydrology 391, 202–216 (2010).
Guttman, N. B. On the Sensitivity of Sample L Moments to Sample Size. Journal of Climate 7, 1026–1029 (1994).
Kumar, R., Samaniego, L. & Attinger, S. Implications of distributed hydrologic model parameterization on water fluxes at multiple scales and locations. Water Resources Research 49, 360–379 (2013).
Samaniego, L., Kumar, R. & Attinger, S. Multiscale parameter regionalization of a grid-based hydrologic model at the mesoscale. Water Resources Research 46 (2010).
Cornes, R. C., van der Schrier, G., van den Besselaar, E. J. M. & Jones, P. D. An Ensemble Version of the E-OBS Temperature and Precipitation Data Sets. Journal of Geophysical Research: Atmospheres 123, 9391–9409 (2018).
Danielson, J. J. & Gesch, D. B. Global multi-resolution terrain elevation data 2010 (GMTED2010). USGS Numbered Series 2011-1073, U.S. Geological Survey (2011).
Hengl, T. et al. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE 12, e0169748 (2017).
Arino, O. et al. Global land cover map for 2009 (GlobCover 2009). PANGAEA https://doi.org/10.1594/PANGAEA.787668 (2012).
Tucker, C. J. et al. An extended AVHRR 8-km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. International Journal of Remote Sensing 26, 4485–4498 (2005).
Hanel, M. et al. Revisiting the recent European droughts from a long-term perspective. Scientific Reports 8, 9499 (2018).
Samaniego, L. et al. Hydrological forecasts and projections for improved decision-making in the water sector in Europe. Bulletin of the American Meteorological Society 100, 2451–2471 (2019).
Warm Winter 2020 Team and ICOS Ecosystem Thematic Centre. Warm winter 2020 ecosystem eddy covariance flux product for 73 stations in FLUXNET-Archive format— release 2022-1. ICOS Carbon Portal https://doi.org/10.18160/2G60-ZHAK (2022).
Wu, H., Svoboda, M. D., Hayes, M. J., Wilhite, D. A. & Wen, F. Appropriate application of the standardized precipitation index in arid locations and dry seasons. International Journal of Climatology 27, 65–79 (2007).
Cancelliere, A., Mauro, G. D., Bonaccorso, B. & Rossi, G. Drought forecasting using the Standardized Precipitation Index. Water Resources Management 21, 801–819 (2007).
Seiler, R. A., Hayes, M. & Bressan, L. Using the standardized precipitation index for flood risk monitoring. International Journal of Climatology 22, 1365–1376 (2002).
Vicente-Serrano, S. M. & López-Moreno, J. I. Hydrological response to different time scales of climatological drought: An evaluation of the Standardized Precipitation Index in a mountainous Mediterranean basin. Hydrology and Earth System Sciences 9, 523–533 (2005).
WMO. Standardized Precipitation Index User Guide. No. 1090 in WMO-No. (World Meteorological Organization, Geneva, Switzerland, 2012).
Asadi Zarch, M. A., Sivakumar, B. & Sharma, A. Droughts in a warming climate: A global assessment of Standardized precipitation index (SPI) and Reconnaissance drought index (RDI). Journal of Hydrology 526, 183–195 (2015).
Ault, T. R. On the essentials of drought in a changing climate. Science 368, 256–260 (2020).
AghaKouchak, A. A baseline probabilistic drought forecasting framework using standardized soil moisture index: Application to the 2012 United States drought. Hydrology and Earth System Sciences 18, 2485–2492 (2014).
Hao, Z., AghaKouchak, A., Nakhjiri, N. & Farahmand, A. Global integrated drought monitoring and prediction system. Scientific Data 1, 140001 (2014).
Xu, Y., Wang, L., Ross, K. W., Liu, C. & Berry, K. Standardized Soil Moisture Index for Drought Monitoring Based on Soil Moisture Active Passive Observations and 36 Years of North American Land Data Assimilation System Data: A Case Study in the Southeast United States. Remote Sensing 10, 301 (2018).
Zeri, M. et al. Importance of including soil moisture in drought monitoring over the Brazilian semiarid region: An evaluation using the JULES model, in situ observations, and remote sensing. Climate Resilience and Sustainability 1, e7 (2022).
Zink, M. et al. The German drought monitor. Environmental Research Letters 11, 074002 (2016).
Saha, T. R., Shrestha, P. K., Rakovec, O., Thober, S. & Samaniego, L. A drought monitoring tool for South Asia. Environmental Research Letters 16, 054014 (2021).
Samaniego, L. et al. Anthropogenic warming exacerbates European soil moisture droughts. Nature Climate Change 8, 421–426 (2018).
Boeing, F. et al. High-resolution drought simulations and comparison to soil moisture observations in Germany. Hydrology and Earth System Sciences 26, 5137–5161 (2022).
Stagge, J. H., Tallaksen, L. M., Gudmundsson, L., Van Loon, A. F. & Stahl, K. Candidate Distributions for Climatological Drought Indices (SPI and SPEI). International Journal of Climatology 35, 4027–4040 (2015).
Vicente-Serrano, S. M. & Beguería, S. Comment on ‘Candidate distributions for climatological drought indices (SPI and SPEI)’ by James H. Stagge et al. International Journal of Climatology 36, 2120–2131 (2016).
Stagge, J. H., Tallaksen, L. M., Gudmundsson, L., Van Loon, A. F. & Stahl, K. Response to comment on ‘Candidate Distributions for Climatological Drought Indices (SPI and SPEI)’. International Journal of Climatology 36, 2132–2138 (2016).
Sol’áková, T., De Michele, C. & Vezzoli, R. Comparison between Parametric and Nonparametric Approaches for the Calculation of Two Drought Indices: SPI and SSI. Journal of Hydrologic Engineering 19, 04014010 (2014).
Farahmand, A. & AghaKouchak, A. A generalized framework for deriving nonparametric standardized drought indicators. Advances in Water Resources 76, 140–145 (2015).
Kumar, R. et al. Multiscale evaluation of the Standardized Precipitation Index as a groundwater drought indicator. Hydrology and Earth System Sciences 20, 1117–1131 (2016).
Vergni, L., Todisco, F. & Mannocchi, F. Evaluating the uncertainty and reliability of standardized indices. Hydrology Research 48, 701–713 (2016).
Heidenreich, N.-B., Schindler, A. & Sperlich, S. Bandwidth selection for kernel density estimation: A review of fully automatic selectors. AStA Advances in Statistical Analysis 97, 403–433 (2013).
Scott, D. W. & Sain, S. R. Multidimensional Density Estimation. In Handbook of Statistics, vol. 24, 229–261 (Elsevier, 2005).
Abramowitz, M. & Stegun, I. A. Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, ninth dover printing, tenth gpo printing edn (Dover, New York, 1964).
Rajagopalan, B., Lall, U. & Tarboton, D. G. Evaluation of kernel density estimation methods for daily precipitation resampling. Stochastic Hydrology and Hydraulics 11, 523–547 (1997).
Hargreaves, G. H. & Zohrab, A. Samani. Reference Crop Evapotranspiration from Temperature. Applied Engineering in Agriculture 1, 96–99 (1985).
Thornthwaite, C. W. An Approach toward a Rational Classification of Climate. Geographical Review 38, 55–94 (1948).
Allen, R. G., Pereira, L., Raes, D. & Smith, M. Crop evapotranspiration — guidelines for computing crop water requirements. FAO Irrigation and drainage 56, 327 (1998).
Stagge, J. H., et al. (eds.) Hydrology in a Changing World, IAHS-AISH Proceedings and Reports, 367–373 (Copernicus GmbH, 2014).
Makkink, G. F. Testing the Penman formula by means of lysimeters. Journal of the Institution of Water Engineerrs 11, 277–288 (1957).
Pohl, F. et al. Long-term daily hydrometeorological drought indices, soil moisture, and evapotranspiration for ICOS ecosystem sites. Zenodo https://doi.org/10.5281/zenodo.7561854 (2022).
Hofstra, N., Haylock, M., New, M. & Jones, P. D. Testing E-OBS European high-resolution gridded data set of daily precipitation and surface temperature. Journal of Geophysical Research: Atmospheres 114 (2009).
Bandhauer, M. et al. Evaluation of daily precipitation analyses in E-OBS (v19.0e) and ERA5 by comparison to regional high-resolution datasets in European regions. International Journal of Climatology 42, 727–747 (2022).
Mavromatis, T. Drought index evaluation for assessing future wheat production in Greece. International Journal of Climatology 27, 911–924 (2007).
Harrell, F. E. Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis. Springer Series in Statistics (Springer International Publishing, Cham, 2015).
Agnew, C. T. Using the SPI to Identify Drought. Drought Network News (1994–2001) (2000).
Svoboda, M. et al. The Drought Monitor. Drought Mitigation Center Faculty Publications (2002).
Pohl, F. Long-term daily hydrometeorological drought indices, soil moisture, and evapotranspiration for ICOS ecosystem sites: software code. https://doi.org/10.5281/zenodo.7473637 (2023).
Acknowledgements
We acknowledge the E-OBS dataset from the EU-FP6 project UERRA (http://www.uerra.eu) and the Copernicus Climate Change Service, and the data providers in the ECA&D project (https://www.ecad.eu). mHM simulations were conducted at the High-Performance Computing (HPC) Cluster EVE, a joint effort of both the Helmholtz Centre for Environmental Research-UFZ and the German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig. We are grateful to Lily-belle Sweet for her proofreading, which improved the English and quality of the final manuscript.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Author information
Authors and Affiliations
Contributions
F.P. and R.K. designed the study with inputs from O.R., C.R., A.H. and L.S.; O.R. and F.B. provided the simulation data, F.P. calculated the indices and conducted the analysis with support from F.H., F.P. drafted the initial manuscript and all authors provided feedback and contributed to finalizing the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Pohl, F., Rakovec, O., Rebmann, C. et al. Long-term daily hydrometeorological drought indices, soil moisture, and evapotranspiration for ICOS sites. Sci Data 10, 281 (2023). https://doi.org/10.1038/s41597-023-02192-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41597-023-02192-1
This article is cited by
-
Multifractal characterization of meteorological to agricultural drought propagation over India
Scientific Reports (2024)
-
Multi-model hydrological reference dataset over continental Europe and an African basin
Scientific Data (2024)
-
Spatiotemporal analysis of soil moisture drought in the Horn of Africa
Theoretical and Applied Climatology (2024)
-
Observational evidence of legacy effects of the 2018 drought on a mixed deciduous forest in Germany
Scientific Reports (2023)