Hydrologic modeling to examine the influence of the forestry reclamation approach and climate change on mineland hydrology

Highlights

• Conventional-reclamation critically alters how soil water transitions to streamflow.

• Forest-reclamation-approach soil-water environment is similar to unmined soil.

• Reclaimed mine soils were thinner and stored less plant-available water.

• Mineland reclamation & projected climate both provide less water to sustain forests.

Abstract

Forests in the Appalachian region of the U.S. are threatened by a variety of short- and long-term pressures, including climate change, invasive species, and resource extraction. Surface mining for coal is one of the most important drivers of land-use change in the region, reducing native forest cover, causing forest fragmentation, eliminating intact soil, and affecting water resources. The Forestry Reclamation Approach (FRA) has been demonstrated as a successful best practice for restoring forests on mine-impacted landscapes, but little information exists on how the practice will affect hydrologic processes. A study was initiated to examine soil-water movement, as in-situ saturated hydraulic conductivity (K_sat), combined with soil porosity to quantify the potential influence on streamflow of reclaimed mines relative to an unmined, forested control site in eastern Kentucky. We compared different reclamation techniques and time since reclamation to determine the extent to which hydrologic function can be restored. We also simulated evapotranspiration at the watershed scale as a function of reclamation technique for both historical and projected (2050) climate. Results indicate that conventional grassland reclamation critically changes how soil water transitions to streamflow, primarily due to K_sat variability that exceeds that measured for intact and FRA soils. Sites reclaimed using FRA exhibited a soil-water environment that was more similar to the unmined control. However, all reclaimed mine soils were thinner, retained and stored less soil water, and thus could provide less plant-available water during the growing season. The plant-available water stored in reclaimed landscapes may not be sufficient to support forest health and this is exacerbated by projected climate conditions. However, soil development under a combination of FRA techniques has the potential to mitigate this limitation.

Introduction

Headwater stream systems can be substantially affected by surface mining. Since the passage of the Surface Mine Control and Reclamation Act (SMCRA) of 1977 (Wooley, 1978), which requires the completion of a jurisdictional determination on the location, extent, and hydrologic classification of waterways that will be affected by any proposed mining activity, surface coal mines in the Appalachian region in the eastern U.S. have largely been reclaimed as flat, herbaceous areas (Miller and Zégre, 2014), with post-mining land uses of hayland pasture and wildlife habitat, instead of the forested land that existed prior to mining. Surface mining, including mountain-top mining with valley fill, is the cause of the largest area of deforestation in Central Appalachia, affecting an estimated 11,000 km², including 2000–4000 km of stream length (Bernhardt and Palmer, 2011; Miller and Zégre, 2014). Movement of earth by this practice is the most spatially concentrated (approximately 1 gigaton per year) in the U.S. (Hooke, 1999) and an estimated 40 million hectares of land are affected by mining activities around the world (Hooke and Martin-Duque, 2012). This approach to mining removes coal and leaves behind newly exposed and disaggregated parent material (spoils) that are used to mechanically re-structure the surface to something approaching the original shape of the landscape. Due to the expansion of the consolidated rock from blasting, several hundred thousand cubic meters of spoil can end up in the lower part of the stream valley, substantially affecting aquatic and terrestrial ecosystems (Evans et al., 2015; Hester et al., 2019; Pond et al., 2008).

Many components of the natural forest's hydrologic cycle are altered in reclaimed grasslands and shrublands due to the lack of canopy cover and changes to physical properties of the soil (Miller and Zégre, 2014). Soils at conventionally reclaimed grassland sites tend to be compacted by design, thus limiting infiltration and rooting volume and the associated soil structure necessary for water movement in the subsurface (Evans et al., 2015). Water movement through this new material varies at the surface as a function of how much compaction is done, with water traveling 10 m on the order of hours to months (Hester et al., 2019). For example, restoration of surface mined areas in Ohio showed that areas reclaimed as grassland resulted in approximately one-quarter of rainfall becoming stormflow, as opposed to only one-tenth in the undisturbed landscape, although this relation was more varied in the reclaimed basins (Bonta et al., 1997). Similarly, comparison of unmined and reclaimed, surface-mined basins in West Virginia showed elevated stormflow response from the reclaimed basin for events with >2.5 mm h^-1, but decreased streamflow for smaller storms (Messinger, 2003). Stormflow response from reclaimed basins included a rapid runoff response followed by a secondary peak that was attributed to sub-surface flow through valley-fill material. These factors combine to reduce subsurface flow and evapotranspiration, in exchange for increased overland flow, storm response, and potentially increased erosion and sedimentation. However, separating the effect of soil properties from those of a changed vegetation community (Griffith et al., 2012) and landscape structure (Evans et al., 2015) makes it difficult to isolate how the reclamation of the surface material independently affects hydrology.

Restoring headwater stream systems and their ecosystem services requires an understanding of the interconnectedness of hydrologic, geomorphic, and ecological processes (Kauffman et al., 1997). Restoration of aquatic communities occurs when the more basic functions of hydrology, hydraulics, geomorphology and the physiochemical properties have been restored (Harman et al., 2012). This functional approach to stream restoration benefits from the identification and simulation of alternative approaches by focusing on reestablishment of hydrologic functions of precipitation-runoff and providing for base flow, all of which is dependent on surface-water storage and transition of this storage below the root zone (Evans et al., 2015; Fischenich, 2006).

The Forestry Reclamation Approach (FRA) has been shown as an effective means for restoring native forests on lands affected by surface mining in Appalachia, enabling successful reforestation to mixed-mesophytic forests planted as seedlings (Adams, 2017; Burger et al., 2005a). Because spoil compaction is a major impediment of forest restoration success (Angel et al., 2006; Dement et al., 2020), proper spoil handling and placement procedures are emphasized in FRA methods. Strike-off, also known as end-dumped, FRA requires that the mine site be graded to achieve stability, then an approximately 1.5-m deep surface layer of weathered spoils and top soil is placed in piles on top of the stable surface (Sweigard et al., 2005). The piles are leveled with one, and no more than two, passes of a small bulldozer creating an undulated and rough surface. The rough, rocky surface enables rapid infiltration and development of preferential, fast-flow paths that reduce surface runoff and prevent soil erosion (Evans et al., 2015; Hester et al., 2019). Other research indicates that heavy soil compaction on conventionally reclaimed surface mines can be mitigated using a bulldozer with a ripping shank (e.g., Burger et al., 2005b; ripped FRA). The shanks are generally 1-m deep and plowing is implemented in a cross-hatching fashion to maximize soil disturbance (Stram et al., 2017). The resulting landscape is rough, furrowed, and contains many exposed large rocks and boulders. Although ripping achieves a desired physical condition near the surface, the extent of soil fracturing is likely limited below 0.5-m depth. Approximately 6 km² have been reclaimed by FRA (0.05% of the area affected by surface mining in central Appalachia; Angel et al., 2015). To date, FRA has been thoroughly tested for utility in reforestation (Zipper et al., 2011); however, there has been little opportunity to demonstrate the effect of FRA on soil properties and the consequent influence on hillslope and basin hydrology.

With time and forest development, we hypothesize that FRA will provide an environment that functions similar to a natural forest, including soil that promotes infiltration and water storage that can be utilized by tree roots, buffer storm events, and sustain base flow. As such, use of FRA is expected to result in relatively natural patterns in evapotranspiration (ET) that reflect the seasonality that can be absent in reclaimed grasslands. Recent research on stream restoration on mined lands at the University of Kentucky's Guy Cove, an Appalachian basin that was originally reclaimed using standard methods and then converted to FRA using a combination of strike-off and ripped methods, indicates that the FRA can play a vital role in reestablishing storm event hydrology (Blackburn-Lynch, 2015).

Characterization of soil properties resulting from different landscape restructuring and hydrologic simulation as a function of these soil properties provides one way of quantifying the potential benefits of FRA relative to traditional fill-compaction techniques. Soil hydrologic properties under different vegetation have been shown to deviate within decades (Lepilin, 1989; Williamson et al., 2004; Zwieniecki and Newton, 1994). Research in the Appalachian region of Kentucky has used a TOPMODEL-based (Beven and Kirkby, 1979) approach to simulate soil-water storage and streamflow permanence in headwater systems (Williamson et al., 2015a). Related research in western Kentucky evaluated the effect of soil-data sources and different characterizations of soil properties on simulation of soil-water movement and storage along a catena in an agricultural hillslope (Williamson et al., 2014). By replacing base layers of data (land use, soils, topography), this approach can be used for scenario testing. For example, samples from the Natural Resources Conservation Service's (NRCS) National Rapid Carbon Assessment were used to investigate agricultural resiliency for a 30-yr, normalized record centered on 2050 (Baker, 2017). For simulations of the Delaware River Basin, validation of ET estimates for a mixed forest community using a mix of observed and modeled ET (Clark et al., 2012; Senay et al., 2013; Williamson et al., 2015b) improved simulation of water availability given projected climate (Williamson et al., 2016) and were used to simulate the hydrologic effect of the transition of forested areas to agricultural and developed land (Williamson and Claggett, 2019). Simulation of the potential benefits of the FRA will benefit from each of these previous efforts, improving our ability to simulate not only streamflow, but also the water budget that controls seasonal streamflow in forested regions that provide drinking water to large portions of the U.S.

Our objectives included:

• Comparing soil properties (thickness, total porosity, field capacity, available water-holding capacity, and in-situ, field-measured saturated hydraulic conductivity [K_sat]) of reclaimed minelands relative to those for an intact landscape. Characterization of soils that underwent a variety of surface mine reclamation strategies and with different times since reclamation quantify the influence of FRA on soil-water movement and storage, relative to that in intact, unmined hillslopes.

• Simulate the resultant streamflow response as a function of differing soil properties that characterize reclaimed minelands relative to intact soil profiles.

• Model actual evapotranspiration (AET) at the watershed scale as a function of differing soil properties that characterize reclaimed minelands under historical climate conditions and those projected for 2050.