the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jun 2026
Probing the spuriousness of observed downwelling radiative fluxes over the global Tropical Oceans
Abstract. A statistical evaluation of daily downwelling radiation data from the Global Tropical Moored Buoy Array (GTMBA), collocated with Clouds and Earth's Radiant Energy Systems (CERES) satellite data, was conducted from 2000 to 2023. This study addressed systematic biases and spurious data downwelling shortwave (Qs) and longwave (QL) radiation, which are crucial for understanding the ocean's energy budget and climate system. Two filtering methods were applied: a fixed threshold (Qs > 350 Wm-2, QL > 450 Wm-2) and a dynamic threshold (mean + 2 standard deviations per station) to check the spuriousness. For Qs, both methods had a negligible effect or resulted in a slight increase in bias, with no significant improvement in root mean squared error (RMSE) or correlation across regions. However, for QL showing large discrepancies in buoy observations in peaks (420–440 Wm-2) compared to satellite (400–410 Wm-2), both fixed and dynamic threshold filtering consistently improved correlation, bias, and RMSE. This greater effectiveness for QL is attributed to its strong influence by atmospheric temperature and humidity profiles, creating systematic biases that filtering effectively addresses. Overall, threshold filtering proved more effective for QL than for Qs, with fixed methods delivering consistently positive results for QL. The study highlights the need for customised threshold strategies to validate in situ data and advocates for further advancements in satellite retrieval algorithms and data assimilation to enhance the accuracy of radiative flux products and mitigate existing biases.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1946', Anonymous Referee #1, 22 Jun 2026
-
AC1: 'Reply on RC1', kaagita venkatramana, 15 Oct 2026
Detailed Point-by-Point Responses to Referee 1 (RC1)
Reviewer Comment: "The identification of spurious observations requires independent evidence: The manuscript largely treats CERES as a reference and interprets large buoy–satellite discrepancies as spurious buoy measurements. However, disagreement may also arise from CERES retrieval errors or from the representativeness difference between point-scale buoy measurements and gridded satellite products. The authors should validate suspected outliers using additional evidence, such as GTMBA quality-control flags, neighbouring buoy records, instrument metadata, temporal consistency checks, or an independent reanalysis product. The term “potentially spurious observations” would be more appropriate unless the errors can be independently confirmed.
Author Response: We fully agree with the referee's critique regarding the distinction between point buoy observations and 1°x1° satellite spatial grids. We recognize that CERES retrievals serve as a comparative satellite reference rather than absolute ground truth, and discrepancies can stem from spatial representativeness errors (e.g., localized convective cloud cover within a 1° grid cell) or satellite retrieval uncertainties in atmospheric moisture profiles as stated in the limitations section of the manuscript. To address this, we have systematically updated our terminology throughout the manuscript, adopting 'potentially spurious observations' and 'potential spuriousness'. Furthermore, we have expanded our discussion in data and methodology and conclusions to explicitly detail point-to-grid representativeness mismatch alongside physical buoy sensor mechanisms (such as sea-salt accumulation on radiometer domes and bird perching).
Manuscript Revisions Executed: 1. Title Updated: 'Probing the potential spuriousness of observed downwelling radiative fluxes over the global Tropical Oceans' (Page 1, Line 1). 2. Abstract and headings revised: Replaced 'spurious data' with 'potentially spurious observations' across abstract (Page 1, Lines 23-28) and section headings (Page 4, Line 180). 3. Discussion Expanded: Added an explicit representativeness framework in section 4 (Pages 15-16) citing Trolliet et al. (2018) regarding point-to-grid oceanic comparisons.
Reviewer Comment “The fixed and dynamic thresholds require stronger justification.
The fixed thresholds of 350 W m⁻² for Qs and 450 W m⁻² for QL may remove physically plausible daily values under certain atmospheric conditions. Similarly, the mean + 2 standard deviations approach does not fully account for seasonal cycles, geographic differences, or non-normal distributions. The authors should provide sensitivity tests using alternative thresholds and consider station- and season-specific criteria, robust statistics such as the median absolute deviation, or physically based clearsky limits.
Author Response: We thank the reviewer for this valuable comment. We clarify that our dynamic thresholds were already calculated on a station-by-station basis (as tabulated in Tables S1 and S2), which inherently accounts for geographic differences in clear-sky radiation limits. Dynamic thresholds are parameterized station-by-station as Threshold_station = Mean_station + 2 * SD_station using all available daily observations at that specific mooring site. Because Mean_station and SD_station reflect the local climatological clear-sky radiation limits and atmospheric moisture baseline of each site (e.g., hyper-humid Indo-Pacific Warm Pool vs. drier Atlantic trade-wind regions), the threshold adapts dynamically across geographic locations. We have expanded section to detail this parameterization and provided full station-level cutoff values in Supplementary Tables S3 and S4.
To address the reviewer's request for justification and sensitivity evaluation, we performed a comprehensive sensitivity analysis comparing baseline unfiltered data against multiple dynamic cutoffs (Mean ± 1 SD, Mean ± 1.5 SD, and Mean ± 2 SD) alongside fixed cutoffs. In addition, we introduced a dedicated seasonality analysis section (Pages 33–35) and added seasonal envelope distribution plots (Figures 3, 4 and Supplementary Figures S2–S5). Our results demonstrate that station-specific Mean + 2 SD effectively isolates physical extreme spikes without clipping natural clear-sky seasonal cycles.
Manuscript Revisions Executed: 1. Methodology section expanded: Added explicit justification for dynamic threshold parameterization (Page 6, Lines 264-276). 2. New seasonality analysis section: Added Sections ('Seasonality and Threshold Performance', Pages 33-35) with Figures 3-4. 3. Supplementary material updated: Added Tables S1–S4 tabulating station-specific mean, standard deviation, and resulting threshold cutoffs across PIRATA, RAMA, and TRITON.
Reviewer Comment " The construction of daily averages and the treatment of missing data need clarification.
The buoy records contain irregular temporal gaps, which may bias daily means. The authors should specify the minimum number of valid hourly observations required to calculate a daily value, explain how missing hours were handled, and ensure that the temporal definitions of the buoy and CERES daily data are fully consistent. The fraction of filtered observations should also be reported for each station and region. "
Author Response: We appreciate the referee raising this crucial technical detail. To clarify, our study directly ingested pre-aggregated daily mean products provided by NOAA/PMEL (GTMBA) and NASA CERES (SYN1deg Edition 4.1). We did not perform custom hourly-to-daily aggregations, nor did we apply temporal gap-filling or artificial interpolation. GTMBA daily means are pre-calculated by PMEL using an automated 3-stage quality control protocol (Freitag et al., 1999; McPhaden et al., 2023) that requires strict minimum hourly completeness thresholds; days failing coverage criteria are assigned missing values (NaN) and were excluded from our collocated dataset. Both GTMBA and CERES daily means are integrated over synchronized 24-hour UTC calendar days (00:00 to 23:59 UTC).
Manuscript Revisions Executed: 1. Section (Data & Methodology) revised: Added an explicit paragraph detailing PMEL's 3-stage QC pipeline and 24-hr UTC temporal matching (Page 5, Lines 241–245). 2. Main Table 1 updated: Added missing record counts and exact percentages (%) across PIRATA, RAMA, and TRITON arrays. 3. Supplementary Table S2 Added: Tabulated exact missing day counts and percentages per region. While we have also found station-wise missing data, we do not want to present either in manuscript or supplementary information as it would make them clumsy. If you are interested, you can check them from below: https://docs.google.com/spreadsheets/d/1O1VjuvTldkJy2becmpU9Odo-_qNHCX_BqnVXUQlyhQk/edit?usp=drivesdk
https://docs.google.com/spreadsheets/d/1yd-CCqnQavB-PgyhnRz9oCBzOTTS-VEXi1iFSM09-rk/edit?usp=drivesdk
Reviewer Comment " Statistical significance and uncertainty should be quantified.
Some reported improvements in correlation, bias, and RMSE are relatively small. The authors should determine whether these changes are statistically meaningful using paired bootstrap confidence intervals or another appropriate method. Sample sizes should be reported for each threshold category, particularly for the extremevalue subsets, because correlations calculated from a small and restricted range of observations may be unstable. "
Author Response: We agree completely. To rigorously test whether observed metric changes are statistically meaningful, we implemented a non-parametric paired bootstrap algorithm (B = 10,000 resamples). Resampling daily time-index pairs preserves the temporal covariance structure between collocated buoy and satellite observations. For dynamic thresholds, the station threshold (Mean + 2 SD) was re-calculated on each individual bootstrap sample to propagate threshold sampling uncertainty. The results confirm that longwave (QL) filtering produces highly significant error reductions (p < 0.05) across key tropical stations (e.g., RAMA -8°S, 67°E: ΔRMSE = -4.16 W/m², Δr = +0.14, 95% confidence intervals strictly excluding zero). Shortwave (QS) changes, as expected, are smaller and often statistically insignificant.
Manuscript Revisions Executed: 1. Section Added: Introduced 'Paired bootstrap significance testing' protocol (Pages 6–7, Lines 280–312). 2. Main Tables 4–7 added: Tabulated station-level sample sizes (N), filtered record counts, metric changes (ΔRMSE, Δr), 95% CIs, and boolean significance flags (p < 0.05). 3. Results text annotated: Annotated all key metric changes in Section with 95% bootstrap confidence intervals.
Reviewer Comment " The statistical metrics are not consistently described.
The methods section refers to standard deviation, bias, correlation, and RMSE, whereas Equation (1) defines MAE and no explicit equation is provided for bias. The authors should revise the methodology section, define all metrics consistently, and clarify whether improvements refer to absolute bias, signed bias, MAE, or RMSE. "
Author Response: We thank the reviewer for highlighting these inconsistencies. We have thoroughly revised Section to explicitly state the mathematical formulas for all statistical metrics used in this study, ensuring consistent terminology across all main text sections, figures, and supplementary tables.
Manuscript Revisions Executed: 1. Methodology section updated: Added explicit mathematical equations: - MAE (Eq. 1): MAE = (1/n) Σ |y_i - x_i| - Correlation r (Eq. 2): Pearson correlation coefficient formula - RMSE (Eq. 3): RMSE = sqrt((1/n) Σ (y_i - x_i)²) - Standard Deviation SD (Eq. 4): Sample standard deviation formula
Reviewer Comment " The scope of the conclusions should be moderated.
The results show that threshold filtering is generally more effective for QL than for Qs. However, the manuscript does not directly quantify the effects of the identified observations on long-term climate trends, SST variability, or ocean heat-budget calculations. These implications should be presented more cautiously unless additional analyses are provided."
Author Response: We fully agree with the referee. Since our study evaluates surface radiative flux discrepancies without performing long-term heat budget model integrations, claiming direct correction of SST trends overstates our findings. We have moderated our claims across the abstract, discussion, and conclusions sections. Uncorrected radiative flux discrepancies are now framed as 'potential sources of bias in local ocean heat-budget calculations and SST trend analyses that warrant further investigation'.
Manuscript Revisions Executed: 1. Abstract revised: Softened heat budget impact statements (Page 2, Lines 25–28). 2. Discussion section moderated: Reframed Ocean heat budget implications as hypothetical pathways for future modeling studies (Page 14, Lines 603–611). 3. Conclusions section updated: Replaced definitive claims with nuanced recommendations for climate modelers (Pages 14–15).
Reviewer Comment " Minor comments
The manuscript would benefit from careful English editing and a more concise presentation. Several sections are repetitive. The CERES product name, version, spatial resolution, temporal resolution, and variable definitions should be stated explicitly. Figure labels and legends should also be checked carefully; for example, the legend in Figure 1 appears to contain a typographical error. "
Author Response: We thank the reviewer for spotting this typo and requesting explicit product details. We have updated sections with full product specifications and corrected Figure 1.
Manuscript Revisions Executed: 1. CERES specification added: CERES SYN1deg / EBAF Edition 4.1 (1° x 1° spatial grid, daily temporal resolution, adjusted_sfc_sw_down_all_daily and adjusted_sfc_lw_down_all_daily variables; Page 5, Lines 235–236). 2. Figure 1 legend corrected: Changed 'SW & SW' to 'SW & LW' in Figure 1 legend and text caption (Page 6, Line 252).
-
AC1: 'Reply on RC1', kaagita venkatramana, 15 Oct 2026
-
RC2: 'Comment on egusphere-2026-1946', Anonymous Referee #2, 03 Aug 2026
The scientific question addressed in this study is important, and the proposed approach of using satellite observations to identify potential anomalies in station-based radiation measurements has practical value for improving the quality control of long-term radiative flux observations.
1. The fixed threshold method has been briefly introduced, the dynamic threshold method should also be described, such as how to select and adjust parameters to adapt it to this anomaly detection scenario?
2. The improvement obtained by dynamic threshold filtering is more significant for QL than for Qs, Are the identified systematic discrepancies in different part of ocean predominantly positive or negative? Which biases may related to cloud and which biases may related to atmospheric paramters? Such analysis would also help evaluate uncertainties in satellite-derived radiation products.
3. Previous studies have evaluated satellite-derived radiation products using in situ observations over oceans. The authors should clarify the advances of the proposed approach in summary and conlusion section.
Line 22, Clouds and Earth's Radiant Energy Systems (CERES), phrasing here is incorrect, please check it in the article. Clouds and the Earth's Radiant Energy System
Line 30, for QL showing...
Line 374, Helps detect local...
Several grammatical errors and incomplete sentences should be corrected in the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-1946-RC2 -
AC2: 'Reply on RC2', kaagita venkatramana, 15 Oct 2026
Detailed Point-by-Point Responses to Referee 2 (RC2)
Reviewer Comment " The fixed threshold method has been briefly introduced, the dynamic threshold method should also be described, such as how to select and adjust parameters to adapt it to this anomaly detection scenario? "
Author Response: We thank Referee 2 for this question. Dynamic thresholds are parameterized station-by-station as Threshold_station = Mean_station + 2 * SD_station using all available daily observations at that specific mooring site. Because Mean_station and SD_station reflect the local climatological clear-sky radiation limits and atmospheric moisture baseline of each site (e.g., hyper-humid Indo-Pacific Warm Pool vs. drier Atlantic trade-wind regions), the threshold adapts dynamically across geographic locations. We have expanded section to detail this parameterization and provided full station-level cutoff values in Supplementary Tables S3 and S4.
Manuscript Revisions Executed: 1. Methodology section expanded: Added mathematical formulation and physical rationale for station-specific dynamic thresholding. 2. Supplementary Tables S3 and S4 Added: Tabulated exact mean, standard deviation, and resulting cutoff values for QL and QS across all PIRATA, RAMA, and TRITON stations.
Reviewer Comment " The improvement obtained by dynamic threshold filtering is more significant for QL than for Qs, Are the identified systematic discrepancies in different part of ocean predominantly positive or negative? Which biases may related to cloud and which biases may related to atmospheric parameters? Such analysis would also help evaluate uncertainties in satellite-derived radiation products."
Author Response: We appreciate this insightful suggestion. Though the study is not aimed on this part, we have added detailed discussion in manuscript summary section clarifying that downwelling longwave radiation (QL) discrepancies are overwhelmingly positive in buoy records-manifesting as anomalous high-value spikes reaching 420–440 W/m² compared to satellite values of 400–410 W/m².
Manuscript Revisions Executed: 1. Discussion section expanded: Detailing the physical mechanisms.
Reviewer Comment "Previous studies have evaluated satellite-derived radiation products using in situ observations over oceans. The authors should clarify the advances of the proposed approach in summary and conlusion section "
Author Response: We thank the reviewer for highlighting the need to emphasize our study's scientific contribution. Previous satellite evaluation studies (e.g., Zhang et al., 2004; Kato et al., 2018) focused primarily on regional or basin-wide mean bias evaluations, which blend valid data with localized sensor errors and obscure isolated spurious events. Our work advances the field by delivering a station-by-station dynamic anomaly detection framework across all three global tropical buoy networks (RAMA, PIRATA, TRITON). This approach isolates site-specific instrument malfunctions and microclimatic anomalies without corrupting valid regional climate records.
Manuscript Revisions Executed: 1. Introduction section revised: Explicitly articulated the novel station-by-station dynamic screening approach relative to prior basin-mean studies (Pages 3–4). 2. Summary & Conclusions updated: Emphasized actionable recommendations for automated quality control in global ocean observing systems (Pages 14–15).
Reviewer Comment "Correct the definition of CERES to 'Clouds and the Earth's Radiant Energy System'."
Author Response: Corrected the acronym definition in Sections to 'Clouds and the Earth's Radiant Energy System (CERES)'.
Reviewer Comment " Several grammatical errors and incomplete sentences should be corrected in the manuscript"
Author Response: Manuscript is carefully revised and corrected.
-
AC2: 'Reply on RC2', kaagita venkatramana, 15 Oct 2026
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 183 | 0 | 2 | 185 | 0 | 0 |
- HTML: 183
- PDF: 0
- XML: 2
- Total: 185
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript investigates potentially spurious downwelling shortwave (Qs) and longwave (QL) radiation observations from the Global Tropical Moored Buoy Array by comparison with collocated CERES satellite data. The topic is relevant to ocean heat-budget studies and quality control of in situ observations. The analysis covers three major tropical mooring networks and provides potentially useful station-level information. However, the current methodology does not yet fully support the identification of “spurious” buoy observations.
Major comments
1. The identification of spurious observations requires independent evidence.
The manuscript largely treats CERES as a reference and interprets large buoy–satellite discrepancies as spurious buoy measurements. However, disagreement may also arise from CERES retrieval errors or from the representativeness difference between point-scale buoy measurements and gridded satellite products. The authors should validate suspected outliers using additional evidence, such as GTMBA quality-control flags, neighbouring buoy records, instrument metadata, temporal consistency checks, or an independent reanalysis product. The term “potentially spurious observations” would be more appropriate unless the errors can be independently confirmed.
2. The fixed and dynamic thresholds require stronger justification.
The fixed thresholds of 350 W m⁻² for Qs and 450 W m⁻² for QL may remove physically plausible daily values under certain atmospheric conditions. Similarly, the mean + 2 standard deviations approach does not fully account for seasonal cycles, geographic differences, or non-normal distributions. The authors should provide sensitivity tests using alternative thresholds and consider station- and season-specific criteria, robust statistics such as the median absolute deviation, or physically based clear-sky limits.
3. The construction of daily averages and the treatment of missing data need clarification.
The buoy records contain irregular temporal gaps, which may bias daily means. The authors should specify the minimum number of valid hourly observations required to calculate a daily value, explain how missing hours were handled, and ensure that the temporal definitions of the buoy and CERES daily data are fully consistent. The fraction of filtered observations should also be reported for each station and region.
4. Statistical significance and uncertainty should be quantified.
Some reported improvements in correlation, bias, and RMSE are relatively small. The authors should determine whether these changes are statistically meaningful using paired bootstrap confidence intervals or another appropriate method. Sample sizes should be reported for each threshold category, particularly for the extreme-value subsets, because correlations calculated from a small and restricted range of observations may be unstable.
5. The statistical metrics are not consistently described.
The methods section refers to standard deviation, bias, correlation, and RMSE, whereas Equation (1) defines MAE and no explicit equation is provided for bias. The authors should revise the methodology section, define all metrics consistently, and clarify whether improvements refer to absolute bias, signed bias, MAE, or RMSE.
6. The scope of the conclusions should be moderated.
The results show that threshold filtering is generally more effective for QL than for Qs. However, the manuscript does not directly quantify the effects of the identified observations on long-term climate trends, SST variability, or ocean heat-budget calculations. These implications should be presented more cautiously unless additional analyses are provided.
Minor comments
The manuscript would benefit from careful English editing and a more concise presentation. Several sections are repetitive. The CERES product name, version, spatial resolution, temporal resolution, and variable definitions should be stated explicitly. Figure labels and legends should also be checked carefully; for example, the legend in Figure 1 appears to contain a typographical error.