Based On The Spatiotemporal Characteristics Of Chlorophyll A Concentration Monitoring In The Yellow Sea And Bohai Sea

Good sea water resources are an important guarantee for the sustainable development of the marine ecological environment.The increasing eutrophication of water bodies leads to frequent outbreaks of harmful algae.Chlorophyll a can be used as an important indicator for quantitative analysis of the growth of marine algae and other phytoplankton.It is closely related to sea surface temperature,changes in sea surface wind field,wave field and human activities.It explores the spatiotemporal variation and response of chlorophyll a concentration and environmental factors,and uses long short-term memory network（LSTM）to predict chlorophyll a concentration,which has played an irreplaceable role in monitoring harmful blooms.Based on the satellite remote sensing data of the mid-resolution imaging spectrometer（MODIS/Aqua）from 2003 to 2017,the spatiotemporal changes of the long time chlorophyll a concentration series in the Yellow Sea and Bohai Sea were analyzed.The results show that the chlorophyll a concentration in the Yellow Sea and Bohai Sea has a spatially decreasing trend from the coastal area to the deep-water basin.From the point of view of interannual changes,the concentration of chlorophyll a in the Yellow Sea and Bohai Sea showed an overall upward trend in which the growth rate decreased,but there are opposite trends in some areas.The chlorophyll a concentration in the Yellow Sea and Bohai Sea has obvious periodic and seasonal changes,reaching a minimum of2.41 mg·m^-3in July and a maximum of 3.43 mg·m^-3in April.In addition,chlorophyll a concentration in the Laizhou Bay waters was also found the maximum value appears in September.Through the Hurst index（H>0.5）,it was found that there is long memory in the concentration of chlorophyll a.The spatiotemporal changes of chlorophyll a concentration in the Yellow Sea and Bohai Sea are affected by the spatiotemporal changes of sea surface temperature,sea surface wind field and wave field.Through analysis,it is found that there are obvious seasonal changes in sea surface temperature,sea surface wind field and wave field in the Yellow Sea and Bohai Sea..In summer,the southerly wind prevails in the Yellow Sea and Bohai Sea.The temperature is higher than that of other seasons.The temperature difference across the sea is relatively small and there is no obvious boundary.The northerly wind prevails in winter,and the temperature is relatively low.The temperature of the entire sea area gradually decreases from south to north and there are zones with obvious temperature differences.The waves are relatively large.The seasons of spring and autumn are the season when the wind speed and direction of the sea surface wind change,the wind direction is chaotic and the stability is relatively poor.The wind speed and direction of the sea surface wind in the Yellow Sea and Bohai Sea area show relatively strong annual fluctuations.The wave field and wind field in the Yellow Sea and Bohai Sea vary basically in each season.Generalized Additive Model（GAM）was used to analyze the influence of environmental factors（sea surface temperature,wind speed,mean wave direction and significant height of the wave）on the change of chlorophyll a concentration.The results showed that there was a significant non-linear correlation between the concentration of chlorophyll a and sea surface temperature,wind speed,mean wave direction and significant height of the wave.The explanation rate of these environmental factors for the change of chlorophyll a concentration was 47.9%,which had a significant effect on the change of chlorophyll a concentration.Compared with wind speed,mean wave direction and significant height of the wave,sea surface temperature can better explain changes in chlorophyll a concentration in the Yellow Sea and Bohai Sea.In addition,the increase of wind direction and human activities（such as the“Water and Sand Adjustment”project,river discharge）also play an important role in changing the chlorophyll a concentration distribution in the Yellow Sea and Bohai Sea.Chlorophyll a concentration is time-dependent to a certain extent,and has seasonal,periodic changes and long memory changes.The LSTM network can handle this time-series well.Based on the analysis of chlorophyll a concentration time-series in the Yellow Sea and Bohai Sea,this paper establishes an LSTM network model and a Gated Recurrent Unit（GRU）network model based on LSTM deformation to describe the changes of chlorophyll a concentration in the Yellow Sea and Bohai Sea.By establishing Recurrent Neural Network（RNN）,LSTM network model and GRU network model to predict the concentration of chlorophyll a,it is found that the accuracy of RNN is significantly lower than that of LSTM network model and GRU network model,but the running time is the least.The prediction results of the LSTM network model and GRU network model are not much different,but the running time of the GRU network model is shorter.As the number of prediction days increases,the prediction accuracy of the model will gradually decrease,but it can still predict the chlorophyll a concentration in the Yellow Sea and Bohai Sea area relatively well.