基于多源遥感数据的拉森冰架形态演变及表面流速估算研究

发布时间：2018-05-04 22:33

  本文选题：拉森冰架 + 光学影像　； 参考：《南京大学》2016年博士论文

【摘要】：南极冰盖与全球气候、生态环境及人类社会未来发展等重大问题密切相关。在极地地区,气候变暖的现象会被放大,过去半个世纪南极半岛所有季节的大气温度均已出现明显升高,其中最为剧烈的增温出现在冬季。地处南极半岛的Faraday/Vernadsky站的气象数据显示,其年平均气温以每十年0.56℃的速度升高,然而冬季的平均气温却以每十年1.09℃的速度升高。南极半岛是全球三个气候变暖最明显的区域之一冰架、大气、海洋三者之间的相互作用系统是全球气候系统的重要组成部分,同时也是最活跃的部分。冰架面积占南极冰盖总面积的11%,而且南极内陆累积的75%的陆地冰物质通过冰架这种特殊形式的“关口”向海洋输送,冰山崩解和底部融化构成了冰架向海洋输送物质的两种主要方式。此外,冰架的表面物质交换(积雪累积与融化)和底部物质交换(附着冰的冻结与融化等)也会对整个南极冰盖的物质平衡产生一定的影响。本文将重点聚焦在拉森冰架上,系统性地完成了一个较长时间序列的拉森冰架面积、高程及表面流速的监测,在此基础上结合气温、海温、降水量等数据分析了拉森冰架的形态演变对气候变化的响应,探讨了冰架物质的输送机制,并基于现有研究成果对拉森冰架未来的稳定性进行预测。主要研究结果如下：(1)采用多种历史遥感数据源来分析拉森冰架1968-2015年的面积变化,主要包括解密航片(南极单帧数据)和卫星光学遥感影像(Landsat和MODIS)。得到了一个长时间序列的拉森冰架持续崩塌与后退的监测结果。直到1980年代中期,拉森冰架北部还没发生明显的面积变化,然而拉森A冰架自从1986年开始便逐步地发生剧烈的后退。拉森B冰架从1990年代早期也开始步入与拉森A冰架相同的后退模式。拉森冰架北部持续的后退最终演变成几次巨型的崩塌事件,当前,拉森A冰架已经完全消失,拉森B冰架仅存大约2000km2的浮冰残存于Scar Inlet处。相对于拉森北部冰架在近几十年发生剧烈的崩塌与后退,拉森C冰架的表面结构和范围在过去50年比较稳定,并未展现出有关气候变暖驱动的冰架面积退缩的直接证据,其变化遵循的是冰架正常的前缘波动模式。当前拉森C冰架仍然保持着监测初始周期(1960年代)面积的90%以上,而且仍然呈现出表面形态的稳定性。(2)通过对两种雷达高度计重合数据的比测实验以及参考椭球系统的转换,可以有效地联合T/P和RA-2两种测高数据进行表面高程变化分析。根据拉森冰架发生的几次巨型崩塌事件以及不同时期的冰架范围,将整个高程变化监测分成4个不同监测单元及时段(拉森A残留、拉森B前缘区域、拉森B残留以及拉森C冰架)。然后采用共线分析法分别监测以上4个监测单元的高程变化情况,并在此基础上探求近20年拉森冰架表面高程的时空变化趋势。在南极半岛气候变暖的背景下,拉森冰架表面高程日益降低：拉森A残留呈现出表面高程剧烈降低的趋势,1992-2001年表面高程降低速率为-0.45m/yr,监测周期内高程变化波动较大；1992-2001年拉森B冰架前缘区域的年均高程降幅为-0.19m/yr,该区域的表面高程变化曲线没有出现剧烈的波动；拉森B残留的监测周期最长(1992-2010年),观测周期内表面高程降低幅度为-0.07m/yr。出乎意料之外的是,从表面高程变化曲线还可以发现拉森B残留在2006年后高程变化趋于稳定,表面高程降低的趋势有所缓解。2002-2010年拉森C冰架表面高程降低幅度为-0.032m/yr,而且该区域的表面高程变化曲线还体现出波动性较小的特征,这可能与拉森C冰架自1986年以来没有出现毁灭性的崩塌事件有关。(3)基于中等分辨率的MODIS和高分辨率的TM影像,采用COSI-Corr模块(运用IDL语言镶嵌入ENVI软件)估算了1986-2012年拉森C冰架和拉森B残留的冰架表面流速。估算结果显示：拉森冰架表面流速特征符合半岛模式和裂隙(缝)模式,冰流主要向东流入Weddell海。此外,从表面流速的时间变化趋势来看,拉森C冰架的大部分区域体现出持续但是较为温和的流速升高趋势,拉森C冰架表面流速估算区2006-2012年的整体平均流速比1986-1990年大约升高13.7%。受近20年的几次冰架崩塌事件地影响,拉森B残留的表面流速波动更为剧烈,1986-1990年拉森B残留的总体年平均表面流速略低于拉森C冰架,然而经历几次崩塌事件之后,拉森B残留2006-2012年的表面流速比1986-1990年急剧升高了大约32%,此时,其平均表面流速已经超过拉森C冰架。总之,近30年拉森冰架大部分区域的表面流速均出现流速增快的趋势。(4)通过对NCEP/NCAR夏季月平均气温数据地分析发现,20世纪后50年南极半岛体现出持续增温的趋势。然而,出乎意料之外的是,从2000年后南极半岛气温出现略微降温的现象。虽然近几十年Weddell海的表面温度没有出现明显的增温趋势,但是其深水温度却出现显著的升温趋势。通过分析GPCP逐月平均降水量数据发现,近几十年拉森冰架流域范围内年均降水量变化的正负趋势不够显著。综上,影响近几十年拉森冰架表面物质平衡变化趋势最关键的气象要素是温度,而非降水。气温上升导致冰架表面融水增加以及融池范围扩大,形态较为稳定的拉森C冰架的表面高程降低被日益增加的表面融化与再凝结所主导,受海洋的影响几乎可以忽略不计。而持续后退的拉森A冰架和拉森B冰架的表面高程除了受气温升高导致表面融化与再凝结的驱动外,同时还受海温升高导致的冰架底部融化加强的影响。此外,拉森北部冰架的后退甚至消失使得它对内陆冰物质产生的背向应力也随之减少,导致占冰架物质绝大部分的陆地冰向海洋输送的速度大大加快,直接反映为冰架表面高程的降低和表面流速的加快。南极半岛北部冰架的“毁灭性”崩解与消失的主要机制为：近半个世纪来,随着全球气候变暖的加剧,南极半岛北部出现了日益扩大的融池范围以及更多的表面融水。温暖的表面融水填充入已存在的裂缝,并顺着这些裂缝传播,部分表面融水甚至能穿透了整个冰架厚度,与底部融水共同作用,最终导致冰架的崩塌。综上所述,气温及海温变暖导致了南极半岛北部冰架在冰川学及流变学上经历了一系列前所未有的剧烈变化,如拉森冰架北部持续的崩塌与后退、表面高程持续的降低以及表面流速的加快等。通过观察高分辨率的遥感影像就会发现,拉森B残留冰架表面横向裂隙(缝)广泛发育,而且每年夏季融化季节都会出现大面积的表面融池,因此基于融水填充冰架裂缝传播的机制可以预测拉森B残留在未来一百年内存在完全消失的可能性。拉森C冰架虽然近期形态较为稳定,但是其维持稳定的几种相互关联机制已经被打破,因此未来拉森C冰架稳定性的威胁也大为增加。近期(2000年以后)区域性的气温变冷在一定程度上会对冰架的冰川学及流变学产生积极的影响,但是气温略微变冷和冰架局部稳定所具有的统计学意义还尚待进一步论证。
[Abstract]:The Antarctic ice cover is closely related to the global climate, the ecological environment and the future development of human society. In the polar regions, the warming phenomenon will be magnified. In the past half century, the atmospheric temperature of all seasons in Antarctic Peninsula has been significantly increased, of which the most intense temperature increases in winter. It is located in the Faraday/ of Antarctic Peninsula. The meteorological data of the Vernadsky station show that the annual average temperature rises at 0.56 degrees centigrade every ten years, while the average temperature in winter rises at 1.09 degrees centigrade every ten years. Antarctic Peninsula is one of the most obvious regions of the world's three warmer regions. The interaction system between the atmosphere and the three oceans is important for the global climate system. The ice shelf is 11% of the total area of the Antarctic ice cover, and 75% of the land ice material accumulated in the Antarctic is transported to the ocean through the special "gate" of the ice shelf, the disintegration of icebergs and the bottom melting constitute the two main ways of the ice shelves to transport the ocean materials. In addition, the ice shelf The surface material exchange (snow accumulation and melting) and the exchange of the bottom material (the freezing and melting of the attached ice) also have a certain effect on the material balance of the whole Antarctic ice sheet. This paper focuses on the Larsen ice shelf, and systematically completes the monitoring of the area, elevation and surface flow velocity of a long time series of Larsen ice shelves. On the basis of the analysis of the temperature, sea temperature, precipitation and other data, the response of the morphological evolution of Larsen ice shelf to the climate change is analyzed, the transport mechanism of the ice shelf material is discussed, and the future stability of the Larsen ice shelf is predicted based on the existing research results. The main results are as follows: (1) the analysis of Larsen using a variety of historical remote sensing data sources The changes in the area of the ice shelf for 1968-2015 years include decryption aerial (Antarctic single frame data) and satellite optical remote sensing images (Landsat and MODIS). The monitoring results of a long time series of continuous collapse and retreat of the Larsen ice rack have been obtained. Until the mid 1980s, there was no obvious area change in the north of the Larsen ice shelf, but the Larsen A ice shelf The Larsen B ice rack has also stepped into the same backward mode as Larsen's A ice shelf since the early 1990s. The continuing retreat of the northern lalson ice shelf eventually evolved into a number of huge collapse events. At present, the Larsen A ice shelf has been completely lost, and the Larsen B ice shelf remains only about 2000km2 of ice floe residue. Deposited at Scar Inlet, the surface structure and scope of the Larsen C ice shelf has been relatively stable over the past 50 years relative to the severe collapse and retreat of the northern ralson ice shelf over the past few decades. It does not show direct evidence of the shrinking of the ice shelf area driven by climate warming. The change follows the normal front fluctuation pattern of the ice shelf. The C ice rack still maintains more than 90% of the area of the initial period (1960s), and still shows the stability of the surface morphology. (2) the surface elevation change analysis can be effectively combined with the T/P and RA-2 altimetry data through the comparison test of the reclosing data of the two radar altimeter and the conversion of the reference ellipsoid system. According to the giant collapse events of Larsen ice shelf and the range of ice shelves at different times, the whole elevation change monitoring is divided into 4 different monitoring units and periods (Larsen A residue, ralson B front area, Larsen B residue and Larsen C ice frame). Then the height variation of the above 4 monitoring units is monitored by collinear analysis. On the basis of this, the temporal and spatial variation trend of the surface elevation of Larsen ice shelf for the last 20 years is explored. Under the background of Antarctic Peninsula climate warming, the surface elevation of Larsen ice shelf is decreasing day by day: the remnant of Larsen A shows a tendency to decrease the surface elevation violently, the surface elevation reduction rate of 1992-2001 years is -0.45m/yr, and the fluctuation of elevation changes in the monitoring period is more than that in the monitoring period. The annual average elevation of the leading edge region of the 1992-2001 year B ice shelf is -0.19m/yr, and the surface elevation curve of the region has no violent fluctuations; the remnant B remains the longest (1992-2010 years), and the decrease in the surface elevation is beyond the expectation of the surface elevation, from the surface elevation curve. It is found that the elevation of Larsen B remains stable after 2006, and the trend of surface elevation decreases to be relieved by the reduction of the elevation of the surface elevation of the rson C ice shelf of -0.032m/yr for.2002-2010, and the surface elevation curve of the region is also characterized by less volatility, which may not appear since 1986 in the Larsen C ice shelf. Destructive collapse events are related. (3) based on medium resolution MODIS and high resolution TM images, the COSI-Corr module (embedded in the IDL language embedded in ENVI software) is used to estimate the surface velocity of the 1986-2012 year ralon C ice shelf and the remnant ice shelf of Larsen B. The results show that the surface velocity characteristics of the Larsen ice rack are in accordance with the peninsula model and the fissure. The ice flow mainly flows eastward into the Weddell sea. In addition, in terms of the time change trend of the surface flow velocity, most of the Larsen C ice shelf shows a continuous but relatively mild flow rate rising, and the overall average velocity of the 2006-2012 year overall average velocity of 2006-2012 years in the surface velocity estimation area of the Larsen ice shelf is approximately 20 years higher than that of 1986-1990 years. The residual surface velocity fluctuation of Larsen B is more intense, and the total average annual surface velocity of Larsen B remains slightly lower than the Larsen C ice shelf for 1986-1990 years. However, after several collapse events, the 2006-2012 year surface flow velocity of Larsen B remains about 32% higher than that of 1986-1990 years, and the average surface flow is at this time. The speed has exceeded the Larsen C ice frame. In a word, the surface flow velocity in most of the Larsen ice shelf has increased rapidly in the last 30 years. (4) through the analysis of the monthly mean temperature of NCEP/NCAR in summer, Antarctic Peninsula showed a trend of continuous warming in the last 50 years. However, unexpectedly, from 2000 after 2000. The temperature appears slightly cooling. Although the surface temperature of the Weddell sea has no obvious warming trend in the last few decades, its deepwater temperature has a significant warming trend. By analyzing the monthly mean precipitation data of GPCP, the positive and negative trend of the annual average annual precipitation in the range of the Larsen ice shelf basin is not obvious enough in recent decades. To sum up, the most important meteorological factors that affect the surface material balance of the Larsen ice shelf in recent decades are temperature, not precipitation. The rise of temperature leads to the increase of the surface melting water and the expansion of the melting pool. The surface elevation reduction of the relatively stable C ice shelf is dominated by the increasing surface melting and recondensation, and the sea is subject to the sea. The impact of the ocean is almost negligible. The surface elevation of the Larsen A ice rack and the Larsen B ice shelf is influenced by the increase of the surface melting and recondensation, as well as the melting of the bottom of the ice shelf, which is caused by the increase of sea temperature. The back stress produced by the material also decreases, which leads to the speed of transportation of the land ice to the vast majority of the ice shelf, which is directly reflected as the reduction of the surface elevation of the ice shelf and the acceleration of the surface velocity. The main mechanism of the "devastating" disintegration and disappearance of the ice shelves in the northern part of Antarctic Peninsula is for nearly half a century. The increasing warming of the ball climate, the increasing melting pool area and more surface melt water in northern Antarctic Peninsula. Warm surface melting water filled with existing cracks and spread along these cracks. Some surface melting water even penetrated the thickness of the whole ice frame, combined with the melting of the bottom, eventually leading to the collapse of the ice frame. The temperature and the warming of sea temperature have caused a series of unprecedented dramatic changes in Glaciology and rheology in northern Antarctic Peninsula, such as the continuous collapse and retreat in the north of the Larsen ice shelf, the continuous reduction of the surface elevation and the acceleration of surface flow velocity. The transverse fissure (SEW) of the surface of the B residual ice shelf is widely developed, and a large area of surface melting pool will appear every summer in the summer melting season. Therefore, the possibility of completely disappearing of Larsen's B residue in the next one hundred years can be predicted based on the mechanism of the crack propagation of the ice shelves filled with melt water. The Larsen C ice shelf is more stable in recent years, but it is more stable. Some of the stable interrelated mechanisms have been broken, so the threat of the stability of the C ice shelf is also increased in the future. In the near future, the regional temperature cooling will have a positive effect on the Glaciology and rheology of the ice shelf to a certain extent, but the temperature is slightly cold and the local stability of the ice shelf has the statistics. The significance of the study is yet to be further demonstrated.

【学位授予单位】：南京大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P343.6
，

本文编号：1844984