the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Dec 2025
The OpenSat4Weather dataset: Ku-band satellite link data for precipitation monitoring
Abstract. The TV-SAT signals received by the ground antennas of Satellite Microwave Links (SMLs) can be opportunistically used for identifying and quantifying precipitation. Hence, SMLs can serve as low-cost rainfall sensors complementary to conventional instruments. However, a significant challenge for opportunistic sensors, such as SMLs and their terrestrial counterpart, i.e., commercial microwave links (CMLs), stems from potential ownership issues, possibly hindering progress in the development of processing tools and validation studies. This underscores the critical need for open data. While CML open datasets are already available, there are no large SML datasets in public repositories. To fill this gap, we introduce here the OpenSat4Weather dataset, a comprehensive and openly accessible collection of data from 215 SML sensors located in Southern France, covering a five-month period from August to December 2022. The dataset is accessible at https://doi.org/10.5281/zenodo.16530166. OpenSat4Weather also includes concurrent conventional data: 6-minute rainfall depths from 113 operational rain gauges, and radar-based estimates of rainfall intensity along each SML path. The radar data are derived from the gauge-adjusted weather radar product Panthere from Météo-France. Additionally, ERA5 reanalysis data of the 0-degree isotherm height are provided for rain height estimation, which is essential for accurate conversion of the received signal level into rainfall intensity.
In this paper, we overview the OpenSat4Weather dataset. We detail the data preparation process and draw statistics of data availability. Furthermore, we present a descriptive analysis of the dataset, including an assessment of the observed rain characteristics, based on the rain gauges, and of the SML received power, and a comparison between SML and radar data. Finally, we provide examples of disturbances and anomalous patterns encountered on the SML raw data. Our ultimate goal is to promote open research that can help in accelerating the development of SML-based applications. Indeed, enhancing rainfall monitoring capabilities by opportunistic sensors could be beneficial in those areas where conventional networks are scarce.
- Preprint
(5818 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
- RC1: 'Comment on essd-2025-537', Filippo Giannetti, 20 Feb 2026
-
RC2: 'Comment on essd-2025-537', Anonymous Referee #2, 13 Mar 2026
The manuscript by Nebuloni et al. introduces a dataset of Satellite Microwave Link (SML) measurements for opportunistic precipitation monitoring. The 1‑minute received signal level time series, collected in late summer and fall 2022 from 215 SML stations (August–December 2022), are freely available. This is an important aspect, since in many cases opportunistic signals usable for precipitation measurements are not accessible. The dataset is complemented by ancillary data such as ERA5 0°C isothermal height, and reference data such as nearby rain‑gauge measurements and radar‑based specific attenuation (SPE) along SML paths. The data are openly accessible through Zenodo and are compliant with FAIR principles. Overall, the dataset is quite unique and well described, and some advice is provided to potential users who wish to work with it. The manuscript looks good but could be improved with some minor changes. Below are my suggestions.
- Licensing. According to Zenodo, the license is CC BY 4.0, whereas the manuscript reports CC SA 4.0. Furthermore, the “copyright statement” at page 1 does not appear to be fully compliant with these licenses. The authors listed in the Zenodo record also need to be acknowledged.
- While the number of terminals is quite high, the time period covered is limited. Could the authors describe the types of precipitation phenomena included (stratiform/convective rain, snow, hail), especially considering that winter and spring are not represented?
- The discussion of methods for determining rain height is important. However, the uncertainty due to attenuation in the melting layer seems to be neglected, or at least delegated to statistical models. These models are suitable for designing communication links but are more limited when the goal is to provide real‑time precipitation measurements with a time resolution comparable to that of rain gauges. It is not clear to me what is suggested to cope with melting layer attenuation.
- As pointed out, many SML retrieval models assume rainfall intensity to be vertically constant below the melting layer. In convective situations, this assumption does not hold, and SML rain quantification becomes more challenging. Moreover, SMLs may pass through regions where rain is mixed with graupel or small hail, for which the coefficients in relation (1) are uncertain. Could the authors provide, within the dataset, a classification (e.g., convective vs. stratiform) for each measurement minute (or longer time intervals)?
- Radar data (Sections 2.3, 3.2). My understanding is that Panthere radar data represent gridded surface rainfall intensities. Therefore, the radar rain reference obtained by averaging along the link path implicitly assumes rainfall uniformity up to the rain height. If this interpretation is correct, please clarify it in the manuscript.
- Lines 169–170. It seems that the melting-layer height corresponds to the bottom of the melting layer, i.e., what is called rain height somewhere else in the manuscript. Please clarify the text.
Citation: https://doi.org/10.5194/essd-2025-537-RC2 - AC1: 'Comment on essd-2025-537', Roberto Nebuloni, 23 Apr 2026
Status: closed
-
RC1: 'Comment on essd-2025-537', Filippo Giannetti, 20 Feb 2026
General Comments
This manuscript presents OpenSat4Weather, an openly accessible dataset of Ku-band Satellite Microwave Link (SML) measurements for opportunistic precipitation monitoring. The dataset includes 1-minute received signal level (RSL) time series, taken in southeastern France from 215 SML stations from August to December 2022, together with co-located 6-minute rain gauge observations, radar-based rainfall estimates projected along each SML path, and ERA5-derived hourly 0°C isotherm heights. The data are arranged in NetCDF format and are compliant with OpenSense conventions.
The release of a large SML dataset fills a gap in the community, as no other open SML datasets are available. The manuscript provides a clear description of instrumentation, preprocessing, metadata structure, and known signal disturbances. In terms of relevance, documentation, and openness, the dataset meets the core expectations of ESSD.
The manuscript is generally well written and is suitable for publication after minor changes.
Specific comments
- Include a short discussion of representativeness and limitations of the covered period (August–December 2022), since it does not represent a full annual cycle.
- Provide clarification about temporal alignment between 1-minute SML data and 5-minute radar data.
- Provide quantitative assessment of the frequency occurrence of disturbances (temperature effects, wet antenna, saturation, misalignment).
- Line 105. Add some relevant reference about the presence of the melting layer and its effects on microwave satellite signals.
- Lines 130-132. Do anomalous values include also sun transits behind the satellites?
Technical corrections
- Conclusions. Line 389. Correct the sensor number in "RSL data for 251 SML sensors".
- Correct the following typos: line 387 "convectional"; line 249 "hisotherm".
Citation: https://doi.org/10.5194/essd-2025-537-RC1 -
RC2: 'Comment on essd-2025-537', Anonymous Referee #2, 13 Mar 2026
The manuscript by Nebuloni et al. introduces a dataset of Satellite Microwave Link (SML) measurements for opportunistic precipitation monitoring. The 1‑minute received signal level time series, collected in late summer and fall 2022 from 215 SML stations (August–December 2022), are freely available. This is an important aspect, since in many cases opportunistic signals usable for precipitation measurements are not accessible. The dataset is complemented by ancillary data such as ERA5 0°C isothermal height, and reference data such as nearby rain‑gauge measurements and radar‑based specific attenuation (SPE) along SML paths. The data are openly accessible through Zenodo and are compliant with FAIR principles. Overall, the dataset is quite unique and well described, and some advice is provided to potential users who wish to work with it. The manuscript looks good but could be improved with some minor changes. Below are my suggestions.
- Licensing. According to Zenodo, the license is CC BY 4.0, whereas the manuscript reports CC SA 4.0. Furthermore, the “copyright statement” at page 1 does not appear to be fully compliant with these licenses. The authors listed in the Zenodo record also need to be acknowledged.
- While the number of terminals is quite high, the time period covered is limited. Could the authors describe the types of precipitation phenomena included (stratiform/convective rain, snow, hail), especially considering that winter and spring are not represented?
- The discussion of methods for determining rain height is important. However, the uncertainty due to attenuation in the melting layer seems to be neglected, or at least delegated to statistical models. These models are suitable for designing communication links but are more limited when the goal is to provide real‑time precipitation measurements with a time resolution comparable to that of rain gauges. It is not clear to me what is suggested to cope with melting layer attenuation.
- As pointed out, many SML retrieval models assume rainfall intensity to be vertically constant below the melting layer. In convective situations, this assumption does not hold, and SML rain quantification becomes more challenging. Moreover, SMLs may pass through regions where rain is mixed with graupel or small hail, for which the coefficients in relation (1) are uncertain. Could the authors provide, within the dataset, a classification (e.g., convective vs. stratiform) for each measurement minute (or longer time intervals)?
- Radar data (Sections 2.3, 3.2). My understanding is that Panthere radar data represent gridded surface rainfall intensities. Therefore, the radar rain reference obtained by averaging along the link path implicitly assumes rainfall uniformity up to the rain height. If this interpretation is correct, please clarify it in the manuscript.
- Lines 169–170. It seems that the melting-layer height corresponds to the bottom of the melting layer, i.e., what is called rain height somewhere else in the manuscript. Please clarify the text.
Citation: https://doi.org/10.5194/essd-2025-537-RC2 - AC1: 'Comment on essd-2025-537', Roberto Nebuloni, 23 Apr 2026
Data sets
The OpenSat4Weather dataset: Ku-band satellite link data for precipitation monitoring Roberto Nebuloni et al. https://doi.org/10.5281/zenodo.16530166
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 602 | 218 | 48 | 868 | 39 | 41 |
- HTML: 602
- PDF: 218
- XML: 48
- Total: 868
- BibTeX: 39
- EndNote: 41
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General Comments
This manuscript presents OpenSat4Weather, an openly accessible dataset of Ku-band Satellite Microwave Link (SML) measurements for opportunistic precipitation monitoring. The dataset includes 1-minute received signal level (RSL) time series, taken in southeastern France from 215 SML stations from August to December 2022, together with co-located 6-minute rain gauge observations, radar-based rainfall estimates projected along each SML path, and ERA5-derived hourly 0°C isotherm heights. The data are arranged in NetCDF format and are compliant with OpenSense conventions.
The release of a large SML dataset fills a gap in the community, as no other open SML datasets are available. The manuscript provides a clear description of instrumentation, preprocessing, metadata structure, and known signal disturbances. In terms of relevance, documentation, and openness, the dataset meets the core expectations of ESSD.
The manuscript is generally well written and is suitable for publication after minor changes.
Specific comments
Technical corrections