Monitoring water surface elevation (WSE) in dam reservoirs is an important requirement for water resources management, particularly in arid and semi-arid regions. Although in situ measurements provide adequate accuracy, they are not always available because of limited spatial coverage and difficulties in accessing some locations. Satellite observations can therefore provide a suitable approach for large-scale WSE monitoring. The Sentinel-3 altimetry mission enables the generation of regular WSE time series; however, its limited ground-track coverage can restrict its applicability over some water bodies. In contrast, the SWOT mission, based on radar interferometry and wide-swath water-surface measurements, provides denser spatial sampling.

This study aims to evaluate the capability of SWOT point-cloud data for extracting and monitoring WSE in four reservoirs in Iran Karkheh, Shahid Kazemi (Bukan), Doroodzan, and Manjil and to assess their agreement with Sentinel-3 and the processed DAHITI dataset. In addition, for independent assessment, WSE derived from SWOT, Sentinel-3A, and DAHITI over Lake Nipissing, Canada, was compared with in situ measurements.

SWOT data were obtained from the high-resolution point-cloud product (L2_HR_PIXC) for the period from August 2023 to November 2025. Multi-stage quality control and filtering were then applied to retain suitable water-surface points. The median elevation of valid points for each satellite pass was calculated as the WSE for that pass. Sentinel-3 data were processed using SRAL observations and the OCOG retracking algorithm, followed by the application of atmospheric and geophysical corrections.

The performance of the datasets was assessed through comparisons between SWOT and Sentinel-3 over Bukan and Doroodzan, comparisons with DAHITI over Doroodzan, and comparisons with in situ measurements over Lake Nipissing. Because DAHITI is itself derived from multi-mission satellite altimetry data, it was not considered an independent reference and was used only to assess inter-dataset agreement. Agreement was evaluated using RMSE, mean absolute error, coefficient of determination (R²), and mean difference.

The results showed that the number of valid points and the temporal continuity of the WSE time series varied among the reservoirs (Fig. 4 and Table 3). Karkheh had a mean of 8,393 valid points and produced a relatively continuous time series. Manjil, characterized by a narrow and elongated shape and a mean of 5,202 valid points, exhibited the greatest temporal discontinuity and WSE variability. Bukan had the highest point density, with a mean of 14,309 valid points, and its WSE standard deviation was 3.14 m. Doroodzan showed the lowest WSE variability, with a standard deviation of 0.98 m.

Over Lake Nipissing, comparison with in situ measurements (Fig. 3) showed that DAHITI had the highest agreement with the in situ data, with an RMSE of 0.27 m, an R² of 0.94, and a mean difference of 0.21 m. For SWOT, the RMSE, R², and mean difference were 0.43 m, 0.69, and 0.26 m, respectively. Sentinel-3A yielded an RMSE of 0.91 m and a mean difference of 0.90 m.

For Bukan, the R², RMSE, and mean difference between SWOT and Sentinel-3 were 0.95, 1.06 m, and 0.75 m, respectively. For Doroodzan, the corresponding values were 0.45, 1.10 m, and 0.23 m. Comparison with DAHITI over Doroodzan also indicated greater agreement between SWOT and DAHITI during the study period (Fig. 7 and Table 5). The RMSE and R² were 0.41 m and 0.77 for SWOT and 0.92 m and 0.62 for Sentinel-3A, respectively. However, because DAHITI is itself based on satellite altimetry data, this comparison represents only relative agreement among the datasets and cannot be interpreted as an assessment of the absolute accuracy of SWOT or as definitive evidence of its superiority over Sentinel-3.

Overall, the results demonstrate that SWOT point-cloud data have good potential for extracting and monitoring WSE in the studied reservoirs; however, their performance depends on reservoir geometry and the quality and number of valid points. This dependence was more evident in narrow and elongated reservoirs. The Lake Nipissing assessment further highlights the importance of independent in situ measurements for evaluating the accuracy of altimetry-derived products. Comparisons with DAHITI should likewise be interpreted as assessments of relative agreement rather than independent validation. Longer observation periods, a larger number of reservoirs with diverse geometric characteristics, independent in situ measurements, and more advanced point-cloud filtering approaches could support a more comprehensive assessment of SWOT performance in future studies.