We used a time series of 15 Landsat images from 1976-2008 to map surface mining and mine reclamation for a 25,000 km² area of southern West Virginia, adjacent Kentucky and Virginia where mountain-top removal/valley-fill coal mining is currently most intense. Despite laws requiring reclamation by a combination of re-contouring, seeding and/or planting, most reclaimed surface mines have highly compacted soils with infiltration rates that are at least an order of magnitude lower than unmined areas, meaning that the effects of surface mining are best understood in terms of cumulative mined area (active + reclaimed mines).   
Results/Conclusions

At regional scales, the proportion of actively mined land in any given year is small (0.5 – 1.5% of the total area), but over 10% of our study area has been mined since 1976.   As of 2008, 9.4% is reclaimed mineland, while 1.3% is either an active or unreclaimed mine. The rate of conversion to mineland remained fairly constant over the period of our study. Approximately 0.2% of the land area was converted to mineland each year.  However, some watersheds (e.g., the Mud River) have experienced considerably higher rates of mining (0.5% per year) with mine conversion rates peaking in the early 2000’s.  The effects of mining are most notable when the landscape is evaluated at the watershed scale.  For some HUC 12-digit watersheds (drainage areas of 75-400 km²), cumulative mined area covers as much as 35% of the watershed, and active mined area ranges from 12 to15%.  16 of the 73 mid-sized watersheds we analyzed are over 10% mined.  Patterns of mining are highly aggregated, so at the scale of smaller watersheds (< 50 km²), the region has many watersheds with >50% cumulative mining, which puts communities downstream of heavily mined watersheds at increased risk for flooding.  Our analysis of the frequency distribution of mined area by watershed size shows that highly mined watersheds are considerably smaller than USGS-gaged watersheds, which means that the effects of mining extent in the study area may not be reflected in the existing hydrologic record. Previous studies of the hydrologic impacts of surface mining have shown that the risk of flooding increases as the percent of mining within a watershed increases.  Thus, we infer from these studies that watersheds with >50% mining will be subjected to dramatically elevated risks of flooding.  We present these results in the context of regionwide patterns of recent mountain-top removal/valley-fill mining.