the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Apr 2025
Development of HYPER-P: HYdroclimatic PERformance-enhanced Precipitation at 1 km/daily over the Europe-Mediterranean region from 2007 to 2022
Abstract. Accurate precipitation estimates are essential for a wide range of applications, including climate research, water resource management, agriculture, and natural hazard assessment. However, developing high-quality, long-term daily datasets at fine spatial resolutions remains challenging due to the inherent variability and heterogeneity of precipitation patterns. This study introduces the HYdroclimatic PERformance-enhanced Precipitation (HYPER-P) product, covering Europe and part of the Mediterranean basin from 2007 to 2022 at a 1 km daily resolution. HYPER-P is derived by downscaling and merging multiple data sources, including remote sensing products from Top-Down (TD) and Bottom-Up (BU) approaches, reanalysis datasets, and gridded in situ observations. The downscaling leverages on CHELSA climatology data, while the merging is obtained through a weighted average approach informed by Triple Collocation Analysis.
Four merged products were developed based on multiple combinations of satellite products, observation and reanalysis datasets. The evaluation of these products was conducted through high-resolution validation in three Mediterranean regions with dense observational networks and coarse-resolution validation across Europe and a portion of North Africa. Results indicate that the combination of TD and BU satellite approaches enhance precipitation estimates, with merged products outperforming the parent datasets, especially in regions with sparse gauge coverage. The inclusion of ERA5-Land further improves accuracy over areas characterized by complex topography. The merging of satellite products, particularly the one including ERA5-Land, shows overall strong performance, although challenges remain in validating precipitation estimates where ground observations are limited. This work contributes to advancing precipitation monitoring capabilities, offering valuable tools for scientific and operational applications across Europe and beyond.
- Preprint
(4898 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2025-156', Anonymous Referee #1, 09 May 2025
I reviewed the manuscript "Development of HYPER-P: HYdroclimatic PERformance-enhanced Precipitation at 1 km/daily over the Europe-Mediterranean region from 2007 to 2022". A large number of (satellite) datasets is used aiming to identify the best precipitation dataset for (regions in) Europe and the Mediterranean region. Merged precipitation datasets outperform other datasets. The number of source and reference datasets is extensive and a detailed description of results is provided. It seems quite surprising that the merged datasets using rain gauge data, M1 and M2, are not the best performing datasets.
Major comments
1) L. 93-95: Not much attention is given to radar-based datasets, although some of the employed local reference datasets contain radar data. Indeed, the radar network across Europe is heterogeneous. I also agree that radars "are prone to errors", or better "sources of errors", but satellite products probably have more issues (as confirmed by the description of satellite limitations in the introduction). Not only in terms of accuracy, but also spatial and temporal resolution generally being lower compared to radar data (the latter being less relevant at the daily timescale). Ground-based radar seems a bit underrated in your study, whereas national meteorological and hydrological services typically employ ground-based radar and rain gauge data, sometimes using satellite data to remove non-meteorological echoes. Indeed, rain gauge data are typically needed to improve the radar precipitation estimates to a high enough level. Radars provide a more direct precipitation retrieval compared to most satellite sensors (radiometers, VIS, IR). On the other hand, satellites are often the only available source of information, apart from sparse rain gauge networks that sometimes only provide near real-time information at best, and hence definitely valuable also because long archives are available. Probably, combining radar, satellite and other datasets is the ultimate way forward, especially for areas without ground-based radar coverage or at far range from ground-based radars and rain gauges, where satellites can act as gap fillers. Though the presented study is valuable to assess the performance of merged satellite-based datasets, which is also relevant for areas outside Europe (with similar climates) with sparse radar and rain gauge coverage, the importance of ground-based radars (merged with rain gauge data) for precipitation monitoring across Europe should be emphasized given its accuracy and generally higher spatiotemporal resolution compared to satellite data. Especially some outlook and recommendations concerning the combined use of rain gauge, ground-based radar, and satellite data (and possibly reanalysis data) would be appropriate. Note that several countries have publicly available climatological or real-time gauge-adjusted radar precipitation datasets (e.g., Germany, the Netherlands). And more high-value datasets across Europe are expected to become available in the future. At the European level, unadjusted OPERA radar products (not yet publicly available, but available for research purposes), or even OPERA-based gauge-adjusted radar products exist: a (near) real-time gauge-adjusted radar product (not yet publicly available; Park et al., 2019), a gauge-adjusted climatological radar product EURADCLIM (publicly available; Overeem et al., 2023). Such datasets have been compared to other (satellite-based) datasets (van der Plas et al., 2024; Lombardo and Bitting, 2024; Hammoudeh et al., 2025) and such references may even be used to put your results into perspective and add the "radar perspective". Now, readers may think that the best precipitation information over Europe comes from satellite-based precipitation products and that these products should always be used for near real-time or climatological purposes. I expect that gauge-adjusted radar precipitation products will generally give better estimates (but not everywhere and always, especially not in areas with lower rain gauge network densities), although they also have their own issues, and hence more research and development is needed to further improve these radar-based products (at the European scale also by combining with e.g. satellite and reanalysis data). It would also help in this respect to make clear what kind of datasets are useful for which applications. For instance, radar precipitation products are valuable in real-time (hazardous weather warnings, input for nowcasting), whereas several satellite products, and hence HYPER-P, are not available in real-time (near real-time or climatological applications). This would help the reader to understand which datasets are available for which latency and hence application. Finally, I can also understand that limitations in availability of (gauge-adjusted) radar data hinder their actual use for new (satellite-based) products or applications.2) L. 270-274: The downscaling procedure needs more explanation. First, monthly CHELSA values are averaged to obtain standard year estimates. I assume these data are from the same year as the satellite data to which the downscaling will be applied, which is repeated for each year. Or do you mean that a climatological monthly average value is computed for each pixel? In addition, could you clarify this statement: "The monthly averages are then linearly interpolated temporally for each day of the year". First, "standard year estimates" are obtained, I assume these are "the montly averages". But how do you interpolate the monthly average value temporally for each day of the year? Do you just divide by the number of days? And "standard year estimates" seems to average all monthly values, hence possible seasonal variation is not taken into account in the downscaling procedure. And why don't you use the daily CHELSA values, which would avoid this interpolation and seem to be available from the CHELSA portal? Finally, CHELSA data up to and including the year 2019 were downloaded, whereas the merged datasets run till 2022, suggesting that CHELSA is "only" used to derive average precipitation patterns, not necessarily based on data from the same period as the satellite data.
Minor comments
* L. 92-93: "Radar measurements are also increasingly available in Europe, with more than 700 operational weather radars managed by the EUMETNET within the OPERA". In the respective article it is stated that "The operational weather radar network in Europe covers more than 30 countries and contains more than 200 weather radars.". So "700" needs to be replaced by the still representative "200".
* L. 152-153: "and high-altitude regions, such as Scandinavia and parts of the Alps, with cold winters, short summers, and low precipitation": As can be seen in datasets, such as E-OBS, precipitation is generally high in the Alps and very high in parts of Scandinavia (at least Norway).
* Table 1: The (catchment) size could be added or mentioned in the main text. In addition, the number of rain gauges or the gauge network density could be mentioned for the datasets employing rain gauges.
* I miss a map that shows the study areas (all regions).
* Are the employed satellite precipitation datasets completely independent from rain gauge data? For instance, perhaps IMERG-L is slightly dependent on rain gauge data (monthly bias adjustment) from a previous period?
* L. 214: "combined with the globally gridded satellite from NOAA": add the exact product name.
* L. 244: "were linearly interpolated each day at midnight UTC": are all the daily values for all datasets in Table 1 from 0-24 UTC? For instance, for E-OBS it is known that gauge measurement intervals differ between networks. Could differences in actual daily measurement intervals affect your results?
* You could consider given normalized metrics to increase comparability of results between regions. For instance, the relative bias (bias divided by the mean rainfall of the reference) and the coefficient of variation of the residuals (standard deviation of the residuals divided by the mean rainfall of the reference).
* L. 458: Please modify this sentence: "BIAS maps (Fig. 7", something went wrong in the description.
* L. 542-543: "The absence of gauge stations in this area limits the reliability of both EMO and E-OBS products, which are forced to estimate precipitation data from sources other than rain-gauges and radar": which sources other than rain-gauges and radar are employed for E-OBS?
* I find the suggestion to use and evaluate the merged precipitation products by computing discharges and comparing these to observed discharges is valuable from the use case perspective (and as pointed out by including references for some of the merged products).
* The choice for IMERG-L, being available in near real-time, suggests that the merged datasets are also relevant for near real-time applications. I suggest to briefly describe the purpose of the presented merged datasets, and especially HYPER-P: for near real-time and climatological applications or for climatological applications only given the latency of HYPER-P? Perhaps I missed it, but could you describe the expected update frequency and latency of HYPER-P?References
* Hammoudeh S, Goergen K, Belleflamme A, Giles JA, Trömel S and Kollet S (2025) Evaluating precipitation products for water resources hydrologic modeling over Germany. Front. Earth Sci. 13:1548557. doi: 10.3389/feart.2025.1548557
* Lombardo, K., and M. Bitting, 2024: A Climatology of Convective Precipitation over Europe. Mon. Wea. Rev., 152, 1555–1585, https://doi.org/10.1175/MWR-D-23-0156.1.
* Overeem, A., van den Besselaar, E., van der Schrier, G., Meirink, J. F., van der Plas, E., and Leijnse, H.: EURADCLIM: the European climatological high-resolution gauge-adjusted radar precipitation dataset, Earth Syst. Sci. Data, 15, 1441–1464, https://doi.org/10.5194/essd-15-1441-2023, 2023.
* Park, S., Berenguer, M., and Sempere-Torres, D.: Long-term analysis of gauge-adjusted radar rainfall accumulations at European scale, J. Hydrol., 573, 768–777, https://doi.org/10.1016/j.jhydrol.2019.03.093, 2019
* van der Plas, E., A. Overeem, J. F. Meirink, H. Leijnse, and L. Bogerd, 2024: Evaluation of IMERG and MSG-CPP Precipitation Estimates over Europe Using EURADCLIM: A Gauge-Adjusted European Composite Radar Dataset. J. Hydrometeor., 25, 1177–1190, https://doi.org/10.1175/JHM-D-23-0184.1.Citation: https://doi.org/10.5194/essd-2025-156-RC1 -
AC1: 'Reply on RC1', Paolo Filippucci, 18 Jun 2025
We thank the reviewer for their valuable comments, which contribute to improve our manuscript quality and readability. We were pleased to review them and clarify our work. Please see below the individual replies to all comments in bold.
I reviewed the manuscript "Development of HYPER-P: HYdroclimatic PERformance-enhanced Precipitation at 1 km/daily over the Europe-Mediterranean region from 2007 to 2022". A large number of (satellite) datasets is used aiming to identify the best precipitation dataset for (regions in) Europe and the Mediterranean region. Merged precipitation datasets outperform other datasets. The number of source and reference datasets is extensive and a detailed description of results is provided. It seems quite surprising that the merged datasets using rain gauge data, M1 and M2, are not the best performing datasets.
We thank the reviewer for their comments. Regarding the results of M1 and M2, this is probably due to the use of CPC as rain gauge product, which is characterized by coarse spatial resolution (0.5°) and limited number of gauge stations (around 17’000 for the full world: while E-OBS has ~14’000 just for Europe). We will clarify this in the text (Line 581 of the original manuscript):
“This is probably due to the native coarse resolution of the CPC dataset used in configuration M1 and M2, as well its relatively low number of included rain gauges”
Indeed, if we had included the EOBS or EMO product, the performance might have improved. However, we would like to maintain the possibility to extend the data to other regions where these products are not available.
Additionally, the two datasets are precious for assessing the quality of the merged product, so their inclusion in it have made impossible to understand its performance.
Major comments
1) L. 93-95: Not much attention is given to radar-based datasets, although some of the employed local reference datasets contain radar data. Indeed, the radar network across Europe is heterogeneous. I also agree that radars "are prone to errors", or better "sources of errors", but satellite products probably have more issues (as confirmed by the description of satellite limitations in the introduction). Not only in terms of accuracy, but also spatial and temporal resolution generally being lower compared to radar data (the latter being less relevant at the daily timescale). Ground-based radar seems a bit underrated in your study, whereas national meteorological and hydrological services typically employ ground-based radar and rain gauge data, sometimes using satellite data to remove non-meteorological echoes. Indeed, rain gauge data are typically needed to improve the radar precipitation estimates to a high enough level. Radars provide a more direct precipitation retrieval compared to most satellite sensors (radiometers, VIS, IR). On the other hand, satellites are often the only available source of information, apart from sparse rain gauge networks that sometimes only provide near real-time information at best, and hence definitely valuable also because long archives are available. Probably, combining radar, satellite and other datasets is the ultimate way forward, especially for areas without ground-based radar coverage or at far range from ground-based radars and rain gauges, where satellites can act as gap fillers.
Though the presented study is valuable to assess the performance of merged satellite-based datasets, which is also relevant for areas outside Europe (with similar climates) with sparse radar and rain gauge coverage, the importance of ground-based radars (merged with rain gauge data) for precipitation monitoring across Europe should be emphasized given its accuracy and generally higher spatiotemporal resolution compared to satellite data. Especially some outlook and recommendations concerning the combined use of rain gauge, ground-based radar, and satellite data (and possibly reanalysis data) would be appropriate. Note that several countries have publicly available climatological or real-time gauge-adjusted radar precipitation datasets (e.g., Germany, the Netherlands). And more high-value datasets across Europe are expected to become available in the future. At the European level, unadjusted OPERA radar products (not yet publicly available, but available for research purposes), or even OPERA-based gauge-adjusted radar products exist: a (near) real-time gauge-adjusted radar product (not yet publicly available; Park et al., 2019), a gauge-adjusted climatological radar product EURADCLIM (publicly available; Overeem et al., 2023). Such datasets have been compared to other (satellite-based) datasets (van der Plas et al., 2024; Lombardo and Bitting, 2024; Hammoudeh et al., 2025) and such references may even be used to put your results into perspective and add the "radar perspective". Now, readers may think that the best precipitation information over Europe comes from satellite-based precipitation products and that these products should always be used for near real-time or climatological purposes. I expect that gauge-adjusted radar precipitation products will generally give better estimates (but not everywhere and always, especially not in areas with lower rain gauge network densities), although they also have their own issues, and hence more research and development is needed to further improve these radar-based products (at the European scale also by combining with e.g. satellite and reanalysis data). It would also help in this respect to make clear what kind of datasets are useful for which applications. For instance, radar precipitation products are valuable in real-time (hazardous weather warnings, input for nowcasting), whereas several satellite products, and hence HYPER-P, are not available in real-time (near real-time or climatological applications). This would help the reader to understand which datasets are available for which latency and hence application. Finally, I can also understand that limitations in availability of (gauge-adjusted) radar data hinder their actual use for new (satellite-based) products or applications.
We completely agree with the reviewer. Indeed, Radar observation are far more accurate than satellite in both spatial and temporal resolution. For this reason, we selected them as benchmark. However, our objective is to create a stable product which can be potentially upscaled to other regions, hence we did not consider the merging with radar measurements. We will underline these concepts and the importance of radar measurements for Europe in the introduction, also highlighting the expected potential latency and applications for HYPER-P (Line 93; 129):
“This technology allows for the collection of reliable precipitation information with high temporal resolution (often in the range of minutes), wide spatial coverage (a single weather radar can cover a circular area with a radius of 100-250 km) and better spatial accuracy compared to rain gauges, which only measure precipitation at their exact location). Radars do not measure rainfall directly; instead, they detect the reflectivity of precipitation particles. Radar measurements are hence often combined with rain gauges, to intercalibrate the measurements and obtain more reliable precipitation estimates. However, also this network is heterogeneous in hardware, signal processing, frequency and scanning strategy, therefore their combination is difficult and prone to errors. Moreover, most of the existing weather radars are mainly located in developed countries (Heistermann et al., 2013).”
“This product is generated by downscaling and merging multiple precipitation datasets from different sources: rain gauges, satellite observations (using both top-down and bottom-up approaches), and reanalysis data. The parent datasets are selected based on criteria such as low latency availability or potential, broad spatial coverage, and high accuracy. As a result, the merged product can be made available globally with relatively short latency—approximately one week. Radar measurements are not included in the merging process due to the lack of a global radar dataset and the limited number of weather radars, particularly in developing countries, but they can be used as a valuable reference.”
2) L. 270-274: The downscaling procedure needs more explanation. First, monthly CHELSA values are averaged to obtain standard year estimates. I assume these data are from the same year as the satellite data to which the downscaling will be applied, which is repeated for each year. Or do you mean that a climatological monthly average value is computed for each pixel? In addition, could you clarify this statement: "The monthly averages are then linearly interpolated temporally for each day of the year". First, "standard year estimates" are obtained, I assume these are "the montly averages". But how do you interpolate the monthly average value temporally for each day of the year? Do you just divide by the number of days? And "standard year estimates" seems to average all monthly values, hence possible seasonal variation is not taken into account in the downscaling procedure. And why don't you use the daily CHELSA values, which would avoid this interpolation and seem to be available from the CHELSA portal? Finally, CHELSA data up to and including the year 2019 were downloaded, whereas the merged datasets run till 2022, suggesting that CHELSA is "only" used to derive average precipitation patterns, not necessarily based on data from the same period as the satellite data.
We thank the reviewer for pointing out the lack of clarity in the description of the downscaling procedure. Since CHELSA dataset is not available in real time, and is only provided up to June 2019, we decided to use a single climatology for the full period: first, monthly aggregates were transformed in daily average by dividing each of them by the number of days in the respective month. Then, CHELSA estimate relative to the same calendar month were averaged to obtain 12 maps, each representative of a different month, thus creating a “standard year”. The average monthly values of the standard year were then attributed to the central day of each month of the study period. A linear interpolation was applied to obtain a distribution estimation for each day of the study period, to avoid step-like patterns after the downscaling. We acknowledge that such strategy does not allow for a precise downscaling of the single storm patterns, due to the absence of concurrent high-resolution data. However, we expect that the high-resolution spatial patterns, even if averaged for each month, are useful to better spatialize coarse resolution information. Additionally, although CHELSA daily estimates are available on the website, they are limited to the year 1980 and were therefore not used. Since concurrent high-resolution data are not available for the full period, the influence of daily patterns on the final downscaled product is expected to be limited. The manuscript will be revised to clarify these aspects (Lines 271; 286).
“Since CHELSA dataset is not available in real-time but only up to June 2019, a single standard year climatology was used for the full period: first, monthly aggregates were converted into average daily precipitation by dividing each of them by the number of days in the corresponding month. Then, CHELSA estimates for the same month across 2000-2019 were averaged to obtain 12 maps, each representative of a different month, thereby producing the standard year estimates (Fig. 1a). The average monthly values of the standard year were then attributed to the central day of each corresponding month of the study period. Linear interpolation was then applied to obtain a daily estimate across the entire study period, thus avoiding step patterns after the downscaling.”
“Although this strategy does not allow a precise downscaling of the single storm pattern, due to the absence of a concurrent high-resolution pattern, it is useful to better spatialize coarse resolution information through the year”
Minor comments
* L. 92-93: "Radar measurements are also increasingly available in Europe, with more than 700 operational weather radars managed by the EUMETNET within the OPERA". In the respective article it is stated that "The operational weather radar network in Europe covers more than 30 countries and contains more than 200 weather radars.". So "700" needs to be replaced by the still representative "200".
We thank the reviewer for noticing the error. It will be corrected.
* L. 152-153: "and high-altitude regions, such as Scandinavia and parts of the Alps, with cold winters, short summers, and low precipitation": As can be seen in datasets, such as E-OBS, precipitation is generally high in the Alps and very high in parts of Scandinavia (at least Norway).We thank the reviewer for pointing it out. Indeed, Alps and Norway are characterized mainly by Dfc and ET Koppen climate, but due to the orographic lift effect, the precipitation is higher than usual. This will be opportunely clarified in the text:
“Subarctic and Polar Climates (Dfc, ET), present in Northern Europe and high-altitude regions, such as Scandinavia and parts of the Alps, with cold winters, short summers, and generally low precipitation, even if the orographic lift in some cases affect the precipitation pattern (Bonacina et al., 1945)”
* Table 1: The (catchment) size could be added or mentioned in the main text. In addition, the number of rain gauges or the gauge network density could be mentioned for the datasets employing rain gauges.
We thank the reviewer for pointing it out. We will add in the text the required information, where available, to the reference datasets description of section 2.2.1
* I miss a map that shows the study areas (all regions).
We have added a map of the selected cumulated precipitation product HYPER-P which clearly show the study area in the text (Figure 16, attached)
Figure 16: Cumulated annual precipitation map of HYPER-P precipitation product, obtained by downscaling and merging precipitation product from reanalysis, satellite TD and satellite BU: specifically, ERA5-Land, IMERG Late Run and SM2RAIN-ASCAT.
* Are the employed satellite precipitation datasets completely independent from rain gauge data? For instance, perhaps IMERG-L is slightly dependent on rain gauge data (monthly bias adjustment) from a previous period?
Thanks for asking: no, we have specifically selected product that do not incorporate rain gauge correction. Starting from version 05, IMERG applies the BIAS correction only in Final Run. The others product were also checked. This will be specified by adding the sentence “The selected version does not use any rain-gauge data” to each of the products in section 2.2.2 (https://gpm.nasa.gov/resources/faq/what-are-differences-between-imerg-early-late-and-final-runs-and-which-should-be-used)
* L. 214: "combined with the globally gridded satellite from NOAA": add the exact product name.
We will update the description. It will state:
“CHIRP combines Thermal Infrared satellite precipitation estimates from the Globally Gridded Satellite (GriSat) and the Climate Prediction Center dataset (CPC TIR) from NOAA to produce the precipitation dataset”
* L. 244: "were linearly interpolated each day at midnight UTC": are all the daily values for all datasets in Table 1 from 0-24 UTC? For instance, for E-OBS it is known that gauge measurement intervals differ between networks. Could differences in actual daily measurement intervals affect your results?
We thank the reviewer for highlighting this issue. We confirm that all the datasets, except for E-OBS, were originally available in UTC time. Indeed, we had overlooked the standard time convention used in E-OBS, given its role as a reference product and the relatively small temporal variation across Europe. However, following the reviewer suggestion, we have re-extracted the data and corrected the interpolation accordingly. As a result, EOBS performance has improved slightly, particularly in the coarse-resolution analysis, where regions with larger temporal discrepancies between standard time and UTC are included. Nevertheless, the overall results remain unchanged. We believe that any residual uncertainties related to the interpolation are minimal.
* You could consider given normalized metrics to increase comparability of results between regions. For instance, the relative bias (bias divided by the mean rainfall of the reference) and the coefficient of variation of the residuals (standard deviation of the residuals divided by the mean rainfall of the reference).We thank the reviewer for their comment. Indeed, the relative indices are useful, but they risk overestimating small errors in arid areas and underestimating large errors in wet areas. Hence, we will add a figure of cumulative precipitation for the full study area (Figure 16) to help the reader comparing the results in different region
* L. 458: Please modify this sentence: "BIAS maps (Fig. 7", something went wrong in the description.
We thank the reviewer for the comment. We will modify the sentence. It will state:
“BIAS maps (Fig. 7) show that all satellite-based precipitation products derived from the TD approach significantly underestimate precipitation in these regions, whereas SM2RAIN-ASCAT tends to overestimate it."
* L. 542-543: "The absence of gauge stations in this area limits the reliability of both EMO and E-OBS products, which are forced to estimate precipitation data from sources other than rain-gauges and radar": which sources other than rain-gauges and radar are employed for E-OBS?
We thank the reviewer for the comment. E-OBS uses topography information together with interpolation technique to better spatialize the precipitation estimates. We will clarify this in the text:
“The absence of gauge stations in this area limits the reliability of both EMO and E-OBS products, which are forced to estimate precipitation data from sources other than rain-gauges and radar (e.g. topography information, Cornes et al., 2018).”
* I find the suggestion to use and evaluate the merged precipitation products by computing discharges and comparing these to observed discharges is valuable from the use case perspective (and as pointed out by including references for some of the merged products).
We thank the reviewer for the comment, which validates our analysis
* The choice for IMERG-L, being available in near real-time, suggests that the merged datasets are also relevant for near real-time applications. I suggest to briefly describe the purpose of the presented merged datasets, and especially HYPER-P: for near real-time and climatological applications or for climatological applications only given the latency of HYPER-P? Perhaps I missed it, but could you describe the expected update frequency and latency of HYPER-P?
The expected latency was pointed out in line 315: “The objective product should be available with a relatively low latency (e.g., a week)”. We will further explain this point in the introduction to clarify it and we will highlight its potential applications:
“The parent datasets are selected based on criteria such as low latency availability or potential, broad spatial coverage, and high accuracy. As a result, the merged product can be made available globally with relatively short latency—approximately one week”
“Due to its potential low latency, HYPER-P can be useful for climatological applications like hydrological modeling, agricultural and drought monitoring or climatological studies.”
Regarding the frequency update, we expect to continue working on this dataset to improve it and extend it for other areas. We will highlight this concept in the conclusions (Line 659)
“Efforts will also aim to extend the dataset both spatially and temporally, especially in areas with sparse rain gauge coverage”
References
* Hammoudeh S, Goergen K, Belleflamme A, Giles JA, Trömel S and Kollet S (2025) Evaluating precipitation products for water resources hydrologic modeling over Germany. Front. Earth Sci. 13:1548557. doi: 10.3389/feart.2025.1548557
* Lombardo, K., and M. Bitting, 2024: A Climatology of Convective Precipitation over Europe. Mon. Wea. Rev., 152, 1555–1585, https://doi.org/10.1175/MWR-D-23-0156.1.
* Overeem, A., van den Besselaar, E., van der Schrier, G., Meirink, J. F., van der Plas, E., and Leijnse, H.: EURADCLIM: the European climatological high-resolution gauge-adjusted radar precipitation dataset, Earth Syst. Sci. Data, 15, 1441–1464, https://doi.org/10.5194/essd-15-1441-2023, 2023.
* Park, S., Berenguer, M., and Sempere-Torres, D.: Long-term analysis of gauge-adjusted radar rainfall accumulations at European scale, J. Hydrol., 573, 768–777, https://doi.org/10.1016/j.jhydrol.2019.03.093, 2019
* van der Plas, E., A. Overeem, J. F. Meirink, H. Leijnse, and L. Bogerd, 2024: Evaluation of IMERG and MSG-CPP Precipitation Estimates over Europe Using EURADCLIM: A Gauge-Adjusted European Composite Radar Dataset. J. Hydrometeor., 25, 1177–1190, https://doi.org/10.1175/JHM-D-23-0184.1.
-
AC1: 'Reply on RC1', Paolo Filippucci, 18 Jun 2025
-
RC2: 'Comment on essd-2025-156', Anonymous Referee #2, 28 May 2025
I have reviewed the manuscript 'Development of HYPER-P: HYdroclimatic PERformance-enhanced Precipitation at 1 km/daily over the Europe-Mediterranean region from 2007 to 2022'. I found the datasets well described and useful, including analysis and discussion on performance, uncertainties and applications. I have no further comments.
Citation: https://doi.org/10.5194/essd-2025-156-RC2 -
AC2: 'Reply on RC2', Paolo Filippucci, 18 Jun 2025
We thank the reviewer for their comment, which validates our analysis
Citation: https://doi.org/10.5194/essd-2025-156-AC2
-
AC2: 'Reply on RC2', Paolo Filippucci, 18 Jun 2025
Data sets
4DMED precipitation product: 1 km merged precipitation from CPC, IMERG-LR for mediterranean basin P. Filippucci et al. https://doi.org/10.5281/zenodo.15025397
4DMED precipitation product: 1 km merged precipitation from CPC, IMERG-LR and SM2RAIN-ASCAT for mediterranean basin P. Filippucci et al. https://doi.org/10.5281/zenodo.10402392
4DHydro precipitation product: 1 km merged precipitation from IMERG-LR and SM2RAIN-ASCAT for Europe P. Filippucci et al. https://doi.org/10.5281/zenodo.15025462
4DHydro precipitation product: 1 km merged precipitation from ERA5-Land, IMERG-LR and SM2RAIN-ASCAT for Europe/HYPER-P: HYdroclimatic PERformance-enhanced Precipitation at 1 km/daily over the Europe region from 2007 to 2022 P. Filippucci et al. https://doi.org/10.5281/zenodo.15025514
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 406 | 55 | 20 | 481 | 18 | 23 |
- HTML: 406
- PDF: 55
- XML: 20
- Total: 481
- BibTeX: 18
- EndNote: 23
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1