ERROR IN SST PRODUCT: PROPAGATION IN THE ESTIMATION OF SEA-AIR CO_2 FLUX
DOU Wen-Jie;JIANG Jin-Gang;ZHOU Bin;YU Zhi-Feng;BAI Yan;HE Xian-Qiang;Institute of Remote Sensing and Earth Sciences,Hangzhou Normal University;Key Laboratory of Zhejiang Urban Wetland and Regional Change Research,State Key Laboratory of Satellites Ocean Environment Dynamics Second Institute of Oceanography,State Oceanic Administration;
Estimation of sea-air CO2 flux is indispensable for a wide range of research especially for carbon biogeochemical cycles and globe climate change. The distribution and CO2 flux are highly variable in surface seawater and vary over a broad spectrum in time and space scales, and there is considerable interest in the use of satellite remotely sensed data to provide synoptic maps of sea-air CO2 flux. However, a great deal of uncertainties are associated with the current remote sensing products of sea-air CO2 flux due to many error sources, which limits largely its application on decision making. Taking SST, the major impact factor on the estimation of sea-air CO2 flux as an example, we presented in a flowchart how the error propagates during the flux calculation with parameters including gas transfer velocity(k), sea surface CO 2 solubility(S), and the partial pressure of CO2 at sea surface(pCO2sw). In addition, using Monte Carlo simulation, we analyzed the transfer law and the final contribution from the error to understand how SST error affects the flux interaction. The results indicate that under the assumption that remote sensing SST error is ±0.5°C and in normal distribution, the SST error transfer law was exponential distributed in k parameterization, and approximately exponential distributed in S parameterization, while normally distributed in pCO2sw parameterization and exponential distributed in CO2 flux; and when atmospheric CO2 partial pressure was fixed at the value of 370μatm, SST brought an error of ±1.2 mmol·m?2·day?1 to the final result of the flux. These results may provide a basis and reference for analyzing other parameters of remote sensing.