Dimethylsulfide model calibration in the Barents Sea using a genetic algorithm and neural network

• Published in: Environmental Chemistry
• Main Authors: Qu, Bo, Gabric, Albert, Zeng, Meifang, Lu, Zhifeng
• Format: Article in Journal/Newspaper
• Language: English
• Published: CSIRO Publishing 2016
• Subjects: Earth Sciences not elsewhere classified; Arctic; Arctic Ocean; Barents Sea; Global warming
• Online Access: http://hdl.handle.net/10072/123661; https://doi.org/10.1071/EN14264

Global warming of climate is connected to ecosystem change, especially in the polar oceans. Biogenic emissions of dimethylsulfide (DMS) are the main biogenic source of sulfate aerosols to the marine atmosphere and may change in the Arctic, where warming is currently very rapid. Here, we simulate DMS distribution and sea-to-air flux in the Barents Sea (30–40°E and 70–80°N) for the period 2003–05. A genetic algorithm is used to calibrate the key parameters in the DMS model. We use MODIS satellite chlorophyll-a data and regional DMS field data to calibrate the model. Owing to limited DMS observations in the Arctic Ocean, multiple data sources were used and compared. A back-propagation neural network is used for predicting regional DMS based on previous history time series. Parameter sensitivity analysis is done based on DMS flux output. Global climate model forcings for 1 × CO2 to 3 × CO2 conditions are used to force the biogeochemical model under future climate warming (c. 2080). The simulation results show that under tripled CO2, DMS flux would increase 168 to 279 % from autumn through winter and would increase 112 % during ice melting season. DMS flux would increase much more in ice-melt-affected water. The increased DMS flux under 3 × CO2 indicates that regional warming could slow owing to the emission of DMS in the Arctic, if the increase in emissions of anthropogenic greenhouse gases is controlled. No Full Text