English
高级检索

基于多时相SAR的矿区沉陷积水遥感监测方法研究

A Mining-area Radar Water Index (MRWI) for monitoring water accumulation in coal mining subsidence areas with multi-temporal SAR images

    摘要
  • 摘要: 矿产资源开采引发的地表沉陷积水是高潜水位矿区耕地资源破坏的重要原因，也是制约矿山绿色发展的瓶颈之一。面向矿区沉陷积水早期监测识别需求，建立了基于多时相合成孔径雷达(SAR)遥感影像的矿区沉陷积水监测方法，命名为矿区雷达水分指数(Mining-area Radar Water Index, MRWI)。以山东省龙堌镇某井工(地下开采)煤矿区为研究区域，收集并处理了2017年哨兵1号(Sentinel-1)GRD产品IW模式29景10 m空间分辨率的SAR影像数据，进行了沉陷区积水的动态监测及其演变分析。利用查准率P(precision)、查全率R(recall)以及两者的调和平均数F1 这3种定量指标，对比分析了MRWI与光学水体指数(Visible and Shortwave Infrared Drought Index, VSDI)和单时相雷达水体指数(Sentinel-1 Dual-polarized Water Index, SDWI)的监测性能。结果表明：对于存在明显地表水体覆盖区域，MRWI、VSDI以及SDWI均能有效识别出水体范围，其中MRWI与VSDI相比的P、R和F1平均值分别达到了94.45%、94.83%和94.46%；MRWI与SDWI相比，P、R和F1平均值分别达到了99.86%、94.03%和96.85%。MRWI具有较强的沉陷积水早期监测能力，以2号积水区为例，MRWI在2017年6月6日监测到了沉陷积水现象，此时VSDI和SDWI均不能指示这一现象；在2017年7月12日左右，沉陷积水导致地表被水体覆盖，此时MRWI、VSDI和SDWI均能指示出沉陷积水现象。MRWI具备较强的时序监测能力，基于Sentinel-1雷达数据的MRWI能够做到至少每12 d为一期的时序监测，并且能够利用长时序的结果对矿区沉陷区积水变化进行时空演变趋势分析。MRWI可为高潜水位矿区沉陷积水现象的早期监测与治理提供有力技术支撑。

     

    Abstract: The water accumulation induced by mineral resource exploitation and surface subsidence is a significant contributor to arable land degradation in mining areas with shallow groundwater tables, posing a major bottleneck to sustainable development of mines. To address the need for early monitoring and identification of subsidence-induced waterlogging in mining regions, a Mining-area Radar Water Index (MRWI) was developed. A total of 29 SAR images of Sentinel-1 GRD with 10-meter spatial resolution in IW mode during 2017 were collected and processed, and the dynamic monitoring and evolution analysis of water accumulation in the subsidence area were carried out. The monitoring performance of MRWI, Visible and Shortwave Infrared Drought Index (VSDI) and Sentinel-1 Dual-polarized Water Index (SDWI) was compared and analyzed by using three quantitative indexes: P (precision), R (recall) and their harmonic mean F1. The results showed that: MRWI, VSDI and SDWI could effectively identify the range of water bodies in the areas with obvious surface water coverage, and the average values of P, R and F1 of MRWI compared with VSDI reached 94.45%, 94.83% and 94.46%, respectively. Compared with SDWI, the average values of P, R and F1 of MRWI reached 99.86%, 94.03% and 96.85%, respectively. MRWI has a strong ability to detect subsidence and water accumulation in the early stage. Taking the No. 2 water accumulation area as an example, MRWI detected the phenomenon of subsidence water accumulation on June 6, 2017, and neither VSDI nor SDWI can indicate this phenomenon at this time. On or about July 12, 2017, subsidence caused the surface to be covered by water, at which point MRWI, VSDI, and SDWI were all indicative of subsidence. MRWI has strong time series monitoring capabilities. MRWI based on Sentinel-1 radar data can achieve time series monitoring at least once every 12 d, and can use the results of long time series to analyze the temporal and spatial evolution trend of water accumulation in the subsidence area of the mining area. MRWI provides strong technical support for the early monitoring and treatment of subsidence and water accumulation in mining areas with high groundwater level.

     

    • HTML全文
    • 参考文献(37)
    • 相关文章
    • 施引文献
  • 资源附件(0)

/

返回文章
返回

相关链接