论文部分内容阅读
北极海冰对全球气候起着非常重要的调制作用,海冰范围是海冰监测的基本参数。近40年,北极地区持续变暖,北极海冰显著减少,进而引发北极自然环境恶化、北半球极端天气频发、全球海平面上升等一系列环境和气候问题。准确获取北极海冰范围及其演变趋势,确定海冰变化对全球气候系统的响应,是研究和预测全球气候变化趋势的关键之一。Has ISST和OISST海冰数据集在海冰监测中应用最为广泛,可为北极地区长时间序列海冰变化研究提供基础数据,但这2套数据集空间分辨率相对较低,应用于北极关键区对中国气候响应研究方面存在很大的局限,为解决这一问题和弥补国内海冰监测微波遥感数据的空白,2011年6月27日,国家卫星气象中心(National Satellite Meteorological Center,NSMC)发布了FY(Fengyun,FY)北极海冰数据集,该数据集利用搭载在FY卫星上的微波成像仪(Microwave Radiation Imager,MWRI)数据,使用Enhance NASA Team算法制作,该算法利用前向辐射传输模型模拟北极地区4种海表类型(海水、新生冰、一年冰和多年冰)在不同大气条件下MWRI辐射亮温,进而得到每种大气条件下0~100%的海冰覆盖度查找表(海冰覆盖度每次增加1%),通过观测值与模拟值的比对得到海冰覆盖度,由该数据集计算得到的北极海冰范围在大部分区域与实际情况相符。该产品虽已进行通道间匹配误差修正和定位精度偏差订正,但由于其搭载的微波成像仪(Microwave Radiation Imager,MWRI)天线长度有限,造成传感器探测到的地物回波信号相对较弱,难以区分海冰和近岸附近的陆地,影响了该数据集的精度和应用。为解决这一问题,本文基于美国冰雪中心(National Snow and Ice Data Center,NSIDC)发布的海冰产品对FY海冰数据集进行优化,NSIDC产品利用判断矩阵对海岸线附近的像元进行识别,并对误差像元进行不同程度的修正,由NSIDC产品计算得到的北极海冰范围与实际情况更为符合。数据集优化大大提高了FY海冰数据集的精度,研究结果表明,优化后FY海冰数据集与NSIDC产品相关系数高达0.9997,且二者日、月、年平均最大海冰范围偏差仅为3.5%、1.9%、0.9%,且FY海冰数据集优化过程对其较好的空间分异特征无明显影响。该数据集可正确地反映北极海冰范围及其变化情况,且海岸线附近海冰的分布情况更准确,可为北极海冰变化研究提供可靠的基础数据。
Arctic sea ice plays a very important modulating role on the global climate. Sea ice range is the basic parameter of sea ice monitoring. In the past 40 years, the Arctic continued to warm and the Arctic sea ice significantly decreased. This in turn caused a series of environmental and climate problems such as worsening Arctic natural environment, extreme weather in the northern hemisphere and rising global sea level. Accurate access to the Arctic sea ice range and its evolution trend to determine the response of sea ice changes to the global climate system is one of the keys to the study and prediction of global climate change trends. HasISST and OISST sea ice data sets are the most widely used in the monitoring of sea ice, which can provide the basic data for the study of long-term sea ice changes in the Arctic. However, these two sets of data sets have relatively low spatial resolution and are used in the Arctic key areas In order to solve this problem and make up the gap of microwave remote sensing data of domestic sea ice monitoring, there is a great limitation on the study of climate response in China. On June 27, 2011, the National Satellite Meteorological Center (NSMC) released FY Fengyun, FY Arctic sea ice dataset, which was generated using Microwave Radiation Imager (MWRI) data on the FY satellite using the Enhance NASA Team algorithm, which simulates a forward radiative transfer model The MWRI radiance of four sea surface types (seawater, newborn ice, one year ice and many years of ice) in the Arctic is bright, and then the sea ice coverage of 0-100% under each atmospheric condition is obtained Ice cover each increase of 1%), through the comparison of observed and simulated values of sea ice coverage obtained by the data set calculated Arctic sea ice range in most areas and the actual situation Match. Although the product has been revised to match the error between channels and to correct the deviation of positioning accuracy, due to the limited MWIR antenna length carried onboard the sensor, the echo signal of the object detected by the sensor is relatively weak and difficult The distinction between sea ice and land near nearshore affected the accuracy and application of the data set. In order to solve this problem, this paper optimizes the FY sea ice data set based on the sea ice products released by the National Snow and Ice Data Center (NSIDC). NSIDC products use the judgment matrix to identify the pixels in the vicinity of the coastline The error pixel correction to varying degrees, calculated by the NSIDC products range of Arctic sea ice more in line with the actual situation. The data set optimization greatly improves the accuracy of the FY sea ice data set. The research results show that the correlation coefficient between the optimized FY sea ice data set and the NSIDC product is as high as 0.9997, and the average daily maximum sea ice range deviation is only 3.5 %, 1.9% and 0.9%, respectively, and the FY sea ice data set optimization process had no significant effect on its better spatial heterogeneity. The dataset can correctly reflect the extent and changes of Arctic sea ice, and the distribution of sea ice near the coastline is more accurate, which can provide reliable basic data for Arctic sea ice change research.