Liu Ge, Li Yun-Mei, Lü Heng, Mu Meng, Lei Shao-Hua, Wen Shuang, Bi Shun, Ding Xiao-Lei
Key Laboratory of Virtual Geographic Environment, Ministry of Education, Nanjing Normal University, Nanjing 210023, China.
Jiangsu Center for Collaboration Innovation in Geographical Information Resource Development and Application, Nanjing 210023, China.
Huan Jing Ke Xue. 2017 Sep 8;38(9):3645-3656. doi: 10.13227/j.hjkx.201702192.
Chlorophyll-a (Chl-a) concentrations are usually measured as the proxy of phytoplankton biomass and used to evaluate the trophic status of inland waters. Based on 49 samples taken from two measurement campaigns in Lake Hongze in 2016, we evaluate the performance of five Chl-a estimation algorithms (including the band ratio, three-band, FLH algorithm, MCI, and UMOC algorithms). The results showed that the UMOC model was the most suitable model for the estimation of Chl-a in Lake Hongze. The mean relative error (MRE) of UMOC was 32.30%, much lower than the band ratio algorithm (75.17%), three-band algorithm (62.44%), FLH algorithm (45.87%), and MCI algorithm (56.95%). The best-performing UMOC model was applied to the atmospherically corrected 689 MERIS images between 2002-2012 and long time series MERIS Chl-a concentration estimation products were acquired. Between 2002 and 2012, the mean Chl-a concentration in Lake Hongze was 19.560 mg·m with substantial spatial and temporal variability. Based on the variability of monthly mean Chl-a concentrations in each pixel, the Lake Hongze waterbody was divided into three water types, Region A, Region B, and Region C. The annual mean Chl-a concentrations of Region B and Region C showed no significant changes, while the concentrations in Region A increased markedly. The analysis of the meteorological factors showed that the fluctuations of the annual mean Chl-a concentrations in Region B and Region C were mainly affected by annual precipitation, suggesting that the Chl-a concentrations of these two regions are dominated by the intensity of the lake flow. The annual mean Chl-a concentrations of Region A showed a strong negative correlation with the annual mean wind speed. The descending trend of the annual wind speed may enhance the eutrophication degree of this region, threatening the safety of the water quality of the South-North Water Transfer Project. The Chl-a concentrations showed a strong positive correlation with the distance from the Huaihe Estuary in the wet season suggesting that the Huaihe River has an obvious inhibitory effect on algal biomass in Lake Hongze during this period.
叶绿素a（Chl-a）浓度通常作为浮游植物生物量的指标进行测量，并用于评估内陆水体的营养状态。基于2016年在洪泽湖两次测量活动中采集的49个样本，我们评估了五种Chl-a估算算法（包括波段比值法、三波段法、FLH算法、MCI算法和UMOC算法）的性能。结果表明，UMOC模型是估算洪泽湖Chl-a最合适的模型。UMOC的平均相对误差（MRE）为32.30%，远低于波段比值算法（75.17%）、三波段算法（62.44%）、FLH算法（45.87%）和MCI算法（56.95%）。将性能最佳的UMOC模型应用于2002 - 2012年经过大气校正的689景MERIS影像，获取了长时间序列的MERIS Chl-a浓度估算产品。2002年至2012年期间，洪泽湖的平均Chl-a浓度为19.560 mg·m ，具有显著的空间和时间变异性。基于每个像素月平均Chl-a浓度的变异性，将洪泽湖水体分为A区、B区和C区三种水型。B区和C区的年平均Chl-a浓度没有显著变化，而A区的浓度显著增加。气象因素分析表明，B区和C区年平均Chl-a浓度的波动主要受年降水量影响，这表明这两个区域的Chl-a浓度受湖水流量强度主导。A区的年平均Chl-a浓度与年平均风速呈强负相关。年风速的下降趋势可能会加剧该区域的富营养化程度，威胁南水北调工程水质安全。在湿季，Chl-a浓度与距淮河河口的距离呈强正相关，这表明淮河在此期间对洪泽湖藻类生物量有明显的抑制作用。
