A 25 km Daily Gridded Dataset of Meteorological Variables and High-Impact Weather Events for New-type Power Systems in China (1980&ndash;2016)

Zhang, Feimin; Bi, Kaixuan; Chen, Xing; Yang, Yi; Yang, Fang; Wang, Chenghai

doi:10.5194/essd-2025-621

Preprints

https://doi.org/10.5194/essd-2025-621

Preprints

23 Feb 2026

| 23 Feb 2026

Status: this preprint is currently under review for the journal ESSD.

A 25 km Daily Gridded Dataset of Meteorological Variables and High-Impact Weather Events for New-type Power Systems in China (1980–2016)

Feimin Zhang, Kaixuan Bi, Xing Chen, Yi Yang, Fang Yang, and Chenghai Wang

Abstract. The new-type power system exhibits pronounced “weather dependency”, wherein high-impact weather events can significantly exacerbate operational security risks. A high-quality gridded dataset that involves both meteorological variables and high-impact weather events is of great significance for new-type power systems. In this study, a spatially adaptive optimal interpolation scheme is developed and applied to generate the China New-type Power Systems Meteorological (CNPS-Met) dataset. The CNPS-Met dataset spans from 1980 to 2016 and covers the entire Chinese mainland, with a daily temporal resolution and a 25 km spatial resolution. It includes eight meteorological variables and eleven high-impact weather events, categorized from generation-side, grid-side and demand-side perspectives relevant to new-type power systems. Validation with existing datasets indicates that the CNPS-Met dataset generally exhibits superior performance in meteorological estimation. Specifically, the estimated mean relative errors for 2-m air temperature, 2-m specific humidity, 10-m wind speed, precipitation and surface pressure averaged over the Chinese mainland could be reduced by 1.7 %–18.5 %, 9.0 %–29.6 %, 1.9 %–8.5 %, 2.7 %–18 % and 4.9 %–5.2 %, respectively. On this basis, a series of high-impact weather events critical to the operation of new-type power system are identified. The spatial distribution of their frequency hotspots and intensity extremes are further analyzed. The CNPS-Met dataset is expected to benefit research and applications at the intersection of meteorology and new-type power systems.

Received: 16 Oct 2025 – Discussion started: 23 Feb 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Feimin Zhang, Kaixuan Bi, Xing Chen, Yi Yang, Fang Yang, and Chenghai Wang

Status: final response (author comments only)

RC1:
'Comment on essd-2025-621', Anonymous Referee #1, 17 Mar 2026
This manuscript presents the China New-type Power Systems Meteorological (CNPS-Met) dataset, a novel 25 km daily gridded product for the Chinese mainland covering 1980-2016. The authors validate their dataset against three existing gridded products (CN05.1, CMFD, CDMet), demonstrating superior performance for most meteorological variables. They further analyze the spatial patterns of these power-system-relevant events. Overall, the topic is highly relevant for ESSD, the methodology is sound and addresses a clear limitation of traditional optimal interpolation (OI) methods, and the dataset fills a critical gap at the intersection of meteorology and energy system. The manuscript is well-structured and the results are presented clearly. I would like to recommend an acceptance after a major revision.

Major Comments
Methodological Justification (Lines 216-236): The spatially adaptive optimal interpolation (OI) scheme proposed in this study is sound, which dynamically adjusts the influence radius based on local station density. While the concept is clear, the paper would benefit from a more explicit demonstration of its impact. For example, how does the resulting field vary spatially across China? This would visually confirm that the algorithm is effectively expanding the radius in data-sparse regions (e.g., Tibetan Plateau) and contracting it in data-dense regions (e.g., Eastern China).

Clarification on the Minimum Station Threshold (Lines 227-236): It seems that the performance of the new OI scheme is also influenced by the chosen parameters (minimum station threshold). However, the manuscript does not specify the actual value of , please explicitly state it. Furthermore, what was the rationale for specifying this threshold?

Verification independence (Lines 163-167): The validation uses CN05.1, CMFD, and CDMet. It is crucial to clarify the degree of independence of these datasets from the observations used to create CNPS-Met. All three validation datasets are themselves interpolated products based on station data. While CNPS-Met demonstrably performs better, the validation is not entirely independent (i.e., it's not a comparison against withheld station data). The authors should validate their results against an independent truth, or explicitly state this limitation and perhaps frame the validation more as a “comparison against existing state-of-the-art gridded products”.

Spatio-temporal Mismatch in Event Definition (Section 2e): This is a critical point. The high-impact weather events are defined using hourly thresholds, but the CNPS-Met dataset has a daily temporal resolution. How are these hourly events identified from a daily dataset? For example, a day with a mean wind speed of 15 m/s could still have an hourly gust exceeding 25 m/s. The manuscript must clearly explain this. Besides, is the “frequency” reported the number of days where the event occurred, or an estimate of the number of hours? This distinction is fundamental for the utility of the dataset.

Uncertainty in the Composite Weather Index (CWI) (Lines 322-330, Eq. 13): The CWI is an interesting metric, but its formulation needs clarification. (1) The term “” is not clear, does this represent the maximum possible value of the variable, or the maximum value observed in the dataset at that location? (2) The notation “” in the piecewise function appears to be a typo. Please clarify the condition(s) for calculating the product.

Interpretation of Wind Speed MREs: In this paper, the MRE is selected to validate the dataset performance. The authors note that the improvement in wind speed is “relatively modest” yet it exhibits the “smallest MREs among all meteorological variables”. This apparent paradox needs clarification. Is it because wind speed is inherently harder to improve (due to its intermittency), but also easier to estimate with low relative error because the values themselves are small? The discussion could be expanded to explain why MRE is a suitable metric for wind speed, and whether other metrics like RMSE might tell a different story.

Analysis of Intensity Extremes (Figure 7): Figure 7 is visually rich but the criteria for defining an "intensity extreme" at the 90% confidence level is not described in the text or caption. How is this confidence level calculated? Is it based on a statistical test of the intensity values at a grid point compared to the surrounding area, or is it simply the 90th percentile of intensity for that event type? Please add a clear explanation to the text or the figure caption.

Relevance of Wind Speed Height for Power Systems: The dataset provides wind speed at 10 m. However, for wind power generation, the variable of interest is typically wind speed at turbine hub height (e.g., 70 m). The thresholds for cut-in and cut-out wind speeds defined in Table 2 are based on operational standards that implicitly refer to hub-height wind speeds. Please clarify whether the wind speed provided in the CNPS-Met dataset is intended to represent hub-height conditions. If so, a detailed explanation of the methodology used to extrapolate from 10 m to hub height is essential.

Clarity on CWI for “Heat and Humid Environment” (Lines 484-489, Fig. 10): The manuscript states that the intensity for this event is based on the CWI, which is dimensionless. However, the definition in Table 2 is simply a threshold. How does the CWI in Eq. 13 apply here? The specific variables (n) and their scaling need to be defined for the HHE event. This is currently unclear.

Minor Comments
Line 56: “Chapter 1 in Xin 2023” – It is unusual to cite a book chapter like this. Please provide a proper citation for the book.

Line 69: “It has at spatial resolution” – Should be “It has a spatial resolution”.

Line 79: “699 ground-based observations” – Are these stations, or individual observation points? Please clarify.

Line 121: “In cases of uneven observational coverage, however, the use of a fixed radius... ” – The sentence beginning here is a comma splice. Consider: “In cases of uneven observational coverage, however, the use of a fixed radius can introduce significant errors...”

Line 144: “meteorological stations are densely distributed” – This is redundant with Fig. 1a. Consider removing the text or referencing the figure more directly.

Equation (6): The formatting of the piecewise function is not clear for me. Please confirm its expression.

Line 249: “referred as” – Should be “referred to as”.

Figure 6 caption: “The concentric circles represent different datasets (from inner to outer: CN05.1, CMFD, CDMet and CNPS- Met.” – The closing parenthesis is missing.

Line 431: “Northern Tibet Plateau” – While commonly used, the standard geographic term is “Tibetan Plateau”, thus, “Northern Tibetan Plateau” is acceptable.

Line 460: “relative weak” – Should be “relatively weak”.

Table 2: The table formatting is broken on pages 15 and 16. This is likely a PDF conversion issue, but ensure the final table is clean. Also, the “Impacts” column entries for extreme temperatures and ice accretion are fragmented and difficult to read. Please rewrite them as complete, coherent sentences.
Citation: https://doi.org/10.5194/essd-2025-621-RC1
CC1:
'Comment on essd-2025-621', Guangwei Li, 17 Mar 2026
This manuscript introduces the China New-type Power Systems Meteorological (CNPS-Met) dataset. Methodologically, the spatially adaptive optimal interpolation scheme is an enhancement of classical data assimilation techniques. The main contribution is the classification of eleven high-impact weather events explicitly linked to the generation-side, grid-side, and demand-side vulnerabilities, which may be useful for power engineering. I have some comments that I hope the authors will consider.
1.The results nearby the complex terrain (e.g., Tibetan Plateau periphery) are not encouraging. Is the error in these regions systematic? A scatter plot of error vs. elevation for these regions would be beneficial to explain the reason.
Sub-region Division (Line 332-333, Fig. 1b): The seven sub-regions are divided “according to the spatial distribution and organizational characteristics of the power grid”. A brief explanation or citation for this specific division would be helpful for readers unfamiliar with China's power grid structure. Why is it more suitable for this analysis?

Dataset Access and Usability (Lines 565-569): The data availability statement provides a DOI. It would be helpful for the reader if the authors could briefly describe in the response to reviews (or in the final manuscript) the structure of the NetCDF files. For example, are all 19 variables in a single file? Are there separate files for different time periods? A small note on the expected data volume would also be practical for potential users.

Temporal Coverage: The dataset ends in 2016, which is already a decade behind. While the authors mention future updates, the current end-date limits the dataset’s applicability for recent power system analyses. Is there a specific reason for this cutoff? A more concrete timeline or plan for extending the dataset to the present day would significantly enhance its value and should be included in the “Future work” section.

Lines 96-97: “the methodology employed in the aforementioned datasets is fundamentally based on spatial interpolation” – This is a bit of an overstatement for CMFD, which the authors themselves describe as integrating remote sensing and reanalysis. Consider rephrasing to “relies heavily on spatial interpolation techniques”.

Line 108: "Hunt et al., 2007" – There is an extra comma before "et al." Please correct to "Hunt et al. 2007" for consistency with other citations.

Equation (2) and (5): The use of superscript T for both transpose and iteration number (as ) is confusing. Please use distinct notation, e.g., for the iteration.

Line 247: “The MRE and RMSE closer to 0” – This phrase is grammatically incomplete. Consider: “Values of MRE and RMSE closer to 0, and R² and EF closer to 1, indicate better estimation performance”.

Table 2 (Impacts column): The impacts for extreme high temperature and ice accretion are fragmented and run together. Please rewrite them as complete sentences for clarity.

The notation “” appears to be a typo. Please clarify the intended condition(s).

Line 332: “seven sub-regions are divided” – This phrasing is awkward. Consider “seven sub-regions are defined” or “the study area is divided into seven sub-regions”.

Line 364: “show the lower variability” – Should be “show lower variability”.

Line 434: “Cut- in wind speed are” – Subject-verb agreement error. Should be “Cut-in wind speed is”.
Citation: https://doi.org/10.5194/essd-2025-621-CC1
CC2: 'Comment on essd-2025-621', Peng Liu, 24 Mar 2026

This manuscript introduces the CNPS-Met dataset, a daily, 25-km gridded product (1980–2016) specifically tailored for China’s new-type power systems. By integrating ground observations from over 2,000 stations with the ERA5 reanalysis using a spatially adaptive Optimal Interpolation (OI) scheme, the authors provide essential meteorological variables alongside 11 high-impact weather events categorized by power system vulnerabilities. The dataset addresses a critical gap in energy-meteorology by aligning atmospheric data with operational needs (generation, grid, and demand). To further enhance the manuscript's clarity and formalize the presentation for readers, the following minor refinements are suggested:
Lines 48–49: Suggest changing "exceed" to "exceeding" for grammatical consistency, and "observations" to "observation stations" for better clarity.

Lines 50–51 & 53: Please consider replacing "ground-based observations" with "ground-based observation stations." This change clarifies that the scarcity refers to the physical distribution of the monitoring network rather than the data points themselves.

Line 58: Consider adding “the” before “increasing frequency” for grammatical correctness.

Line 61: “high-quality” should be hyphenated when used as a compound modifier before “gridded dataset.

Line 62: To enhance readability, you might rephrase "power system-relevant high-impact weather events" to "high-impact weather events relevant to power systems."

Lines 69–75: Suggest changing "over China region" to "across the China region" or "covering the Chinese mainland" for a more idiomatic academic expression. Please correct the grammatical errors in these lines. "It has at spatial resolution" should be changed to "It has a spatial resolution," and the comma after "2020" should be replaced with a period to separate the two independent sentences. For consistency and technical accuracy, please replace "observations" with "observation stations" (e.g., "...approximately 2,400 ground-based observation stations"). This clarifies that the number refers to the monitoring locations, not observation records.

Line 78: Please ensure there are spaces between the numbers and units (e.g., "4 km × 4 km") and consider adding a comma after "2020" for better readability.

Line 89: Please change "high-impact weather event" to " high-impact weather events" to ensure plural agreement with "datasets."

Line 121: "...uniformly distributed, In cases of uneven...". The comma should be replaced with a period to separate the two independent sentences.

Line 141: The list of variables contains redundant conjunctions ("and... and..."). Please rephrase to: "...relative humidity at 2 m, surface pressure, and precipitation."

Line 143: Please provide a brief mention of the specific quality control methods used (e.g., spatial consistency or range checks), which would greatly benefit the readers.

Line 145: Please ensure consistent formatting for figure citations throughout the manuscript. You have used both "Figure 1a" (Line 139) and "Fig. 1a" (Line 145).

Line 155: "...at horizontal resolution of 1° ×1°..." should be "...at a horizontal resolution of 1° ×1°...".
I appreciate the authors' significant efforts and their valuable contribution to the community.

Citation: https://doi.org/10.5194/essd-2025-621-CC2
RC2: 'Comment on essd-2025-621', Anonymous Referee #2, 26 Mar 2026

Please see my review in the supplement.

Citation: https://doi.org/10.5194/essd-2025-621-RC2

Feimin Zhang, Kaixuan Bi, Xing Chen, Yi Yang, Fang Yang, and Chenghai Wang

Data sets

A 25 km Daily Gridded Dataset of Meteorological Variables and High-Impact Weather Events for New-type Power Systems in China Feimin Zhang, Kaixuan Bi, Xing Chen, Fang Yang, Yi Yang, Chenghai Wang, Zijian Zhao, Zhiyuan Ma https://www.doi.org/10.12072/ncdc.nieer.db6972.2025

Feimin Zhang, Kaixuan Bi, Xing Chen, Yi Yang, Fang Yang, and Chenghai Wang

Viewed

Total article views: 362 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
265	78	19	362	28	22

HTML: 265
PDF: 78
XML: 19
Total: 362
BibTeX: 28
EndNote: 22

Views and downloads (calculated since 23 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	103	31	3	137
Mar 2026	143	35	16	194
Apr 2026	19	12	0	31

Cumulative views and downloads (calculated since 23 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	103	31	3	137
Mar 2026	143	35	16	194
Apr 2026	19	12	0	31

Viewed (geographical distribution)

Total article views: 361 (including HTML, PDF, and XML) Thereof 361 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 08 Apr 2026

Short summary

The China New-type Power Systems Meteorological (CNPS-Met) dataset covers mainland China from 1980 to 2016 with a daily resolution and a 25 km grid. It includes eight meteorological variables and eleven high-impact weather events related to power generation, grid transmission, and electricity demand. CNPS-Met outperforms other existing datasets in accuracy for most meteorological variables. We also quantified the frequency hotspots and intensity extremes of high-impact weather events.


Total:	0
HTML:	0
PDF:	0
XML:	0