Tidal state as a source of sampling uncertainty in instantaneous-to-daily scaling of mangrove FAPAR

Abstract. A fixed satellite overpass samples mangrove canopies at different stages of the tidal cycle, but the resulting temporal-scaling uncertainty is rarely quantified. We combined half-hourly eddy-covariance measurements from the Qinglangang Mangrove Reserve with harmonically reconstructed water levels to test whether tidal state covaried with an apparent FAPAR proxy and could bias its representation at overpass time. The proxy was 9.8 % higher during exposed than flooded periods, and changepoint analysis located a site-specific transition near −0.3 m relative to mean sea level. A post-changepoint sensitivity analysis based on state-specific light-use-efficiency estimates had a coefficient of variation of 3.7 %, suggesting that calibration choice alone was unlikely to explain the contrast; however, the proxy was not an independent optical measurement of FAPAR. Across 163 valid days, the median instantaneous-to-daily ratio was 0.83, and applying this site-calibrated ratio reduced in-sample RMSE by 28 % relative to a unity-ratio baseline. Sentinel-2 imagery showed directionally compatible spatial patterns, but elevation uncertainty and spatial autocorrelation limited inference. The analysis identifies tidal phase as a source of sampling uncertainty rather than demonstrating a transferable tide-conditioned correction. Testing such a correction will require optical FAPAR, measured water levels, temporally independent evaluation, and additional mangrove sites.