盘锦湿地净初级生产力时空分布特征

Abstract

主要采用中巴地球资源卫星(CBERS)、合成孔径雷达(SAR)和数字高程模型(DEM)数据,通过主成分变换融合算法、分类回归树CART算法和混合像元分解模型结合神经网络算法,进行了盘锦湿地土地覆盖类型分类。充分考虑湿地生态系统的典型特征,将盐分胁迫因子作为估算湿地耐盐植被净初级生产力(NPP)的环境影响因子之一,构建了基于光能利用率和遥感数据的湿地植被净初级生产力模型。分析了盘锦湿地植被NPP的时空分布特征,并研究了盘锦湿地植被NPP对气温和降水的响应特征。 结果表明:2009年盘锦市植被净初级生产力介于0-1175 gC·m^-2·a^-1之间,平均值为553 gC·m^-2·a^-1。盘锦市植被NPP空间分布规律呈东北向西南逐渐递增的趋势。在湿地植被分类类型中,芦苇的单位面积平均NPP最高,达到1016 gC·m^-2·a^-1。 2004-2009年盘锦植被单位面积平均NPP值在缓慢上升,湿地已呈现缓慢恢复的趋势。总体上气温对盘锦湿地主要植被类型芦苇月平均NPP的影响要强于降水。2004-2009年降水对盘锦地区植被年平均NPP的影响强于气温。;In this paper, in order to improve wetland classification precision, firstly extracted wetland type, then classified other land cover type, by these two steps fulfilled land cover type classification in study area. Using hydrological and topographic parameter data, SAR data, and CBERS data to fusion by adopting principal component transform algorithm, and adopting classification regression tree CART algorithm to construct decision tree, extracted the wetland type in study area. Adopting mixed pixel decomposed model combined with neutral network method, classified other land cover types in study area. After adequately considering wetland ecosystem representative feature, in our paper, constructed wetland vegetation net primary productivity model based on light use efficiency and remote sensing data. Genetic algorithm was adopted to confirm max value of maximum light use efficiency model. Linear method was adopted to calculate FPAR value from NDVI value. Terrestrial ecological model was adopted to calculate temperature stress factor. Simpler evaporating ratio was adopted to calculate moisture stress factor. By establishing quantitative relationship between spectral feature band and land surface parameter, estimated salinity stress factor of wetland salt-resistant vegetation. Spatio-temporal distribution feature of Panjin wetland vegetation NPP was analyzed. According to observed data of temperature and precipitation from weather stations in study area in 2004-2009, separately analyzed correlation relationship between temperature and precipitation with month average NPP of main vegetation type reed in Panjin wetland during growth season, as well as correlation relationship between temperature and precipitation with annual average NPP of Panjin vegetation in 2004-2009, acquired respond trend of wetland vegetation NPP towards temperature and precipitation. Results showed: wetland area in Panjin (including water body, wet meadow, saline seepweed, paddy, reed, and pond, which were classified in this paper) about account for 77.2% of total area of Panjin City. Main vegetation in Panjin wetland coved (including wet meadow, Saline Seepweed, paddy, and reed, which were classified in this paper) about 51.6% area of Panjin City. In 2004-2009, among different land cover types of wetland, the area of water body and wet meadow is decreasing, and the area of paddy, reed, saline seepweed, and pond is obviously increasing. Among these land cover types, the increased area of paddy is the biggest (155.19 km²), followed by saline seepweed (24.18 km²), pond (21.50 km²), and reed (10.75 km²). Slowly recovery trend has appeared in Wetland. In 2009, vegetation net primary productivity in Panjin City is between 0-1175 gC·m^-2·a^-1, and average value is 553 gC·m^-2·a^-1. In general, spatial distribution of vegetation NPP in Panjin City show gradually increased trend from northeast to southwest. The high value region of vegetation NPP mainly concentrate in Shuangtaizi River, and Big Liao River watershed, where reed and paddy widely spreaded. Among different wetland vegetation types, average NPP in unit area of reed is the highest, up to 1016 gC·m^-2·a^-1, thereafter is paddy (464 gC·m^-2·a^-1), saline seepweed (377 gC·m^-2·a^-1), and wet meadow (357 gC·m^-2·a^-1). From 2004 to 2009 average vegetation NPP value in unit area in Panjin region is slowly rising. Among them average vegetation NPP in unit area of Panjin City is the biggest in 2009, up to 553 gC·m^-2·a^-1, and the average NPP value is the smallest(511 gC·m^-2·a^-1)in 2007. In general the influence of temperature to NPP of reed which is the main vegetation type in Panjin wetland is greater than precipitation. With month forward, correlation relationship between NPP with temperature and precipitation is gradually decreasing. In 2004-2009 the influence of precipitation to annual average vegetation NPP in Panjin region is greater than temperature.