Legal forms

A Field-Based Model of the Effects of Landcover Changes on Daytime Summer Temperatures in the North Cascades

A Field-Based Model of the Effects of Landcover Changes on Daytime Summer Temperatures in the North Cascades
of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  137 Physical Geography  , 2010, 31 , 2, pp. 137–155. DOI: 10.2747/0272-3646.31.2.137Copyright © 2010 by Bellwether Publishing, Ltd. All rights reserved. A FIELD-BASED MODEL OF THE EFFECTS OF LANDCOVER CHANGES ON DAYTIME SUMMER TEMPERATURES IN THE NORTH CASCADES Michael A. O’Neal, Lyndsey B. Roth, Brian Hanson, and Daniel J. Leathers Department of GeographyUniversity of DelawareNewark, Delaware 19716  Abstract  : Removal or reduction of forest landcover has been shown to increase dailyhigh temperatures during the summer in clearcut and young, regrowth forests relative tomature or old-growth forests. Although these temperature differences are easily observedin individual stands, extrapolating the net temperature effect to an actively logged terrainis made difficult by the heterogeneous mixture of stand heights and densities. To study theeffect of temperature produced by changes in forest landcover on a regional scale, 13temperature sensors were placed in structurally different stands in a 15 × 24 km area of theGifford Pinchot National Forest on the northwest side of Mount Adams, Washington.These sensors operated for 60 days, from July 20, 2008, through September 17, 2008, insnow-free conditions across a range of elevations. Using field data and a previously devel-oped proxy dataset for stand height and density, near-surface daytime air temperaturemodels were developed for the modern and pre-anthropogenic landscapes. Our modelresults indicate that a 0.7° C average daytime summer temperature difference between thepre-anthropogenic and modern landscapes may be attributed to 20th century reductions inforest cover. [Key words: landcover change, Cascades, climate, glaciers, surface temperature.] INTRODUCTIONFor nearly a century, numerous studies conducted in actively logged terrainshave found that midlatitude mature or old-growth forests experience lower dailymaximum near-surface temperatures during the summer months than adjacentclearcut and young-regrowth sites (Pearson, 1914; Fritts, 1961; Hornbeck, 1970;Raynor, 1971; McCaughey, 1985; Carlson and Groot, 1997; Barg and Edmonds,1999; Heithecker and Halpern, 2006; O'Neal et al., 2009). Although the net dailyor seasonal energy balance that results from reduced forest cover may remainunchanged, the difference between the minimum and maximum daily near-surfacetemperatures, and the averages of near-surface daytime and nighttime temperatures,typically increase in a clearcut forest relative to a forested site (e.g., Hornbeck1970). Such increases in extreme temperature conditions are known to result in aseries of cascading negative impacts on hydrological and biological systems withinor near these altered forested landscapes (Thompson et al., 1998; Stott and Marks,2000; Bourque and Pomeroy, 2001; Hashimoto and Suzuki, 2004). Previous studies of the climatic effects resulting from midlatitude forest-coverchange typically focused on local areas, comparing adjacent stands on the order of a few tens of hectares. The variation of near-surface temperatures among a few dif-ferent stands is often described as a function of the maturity of the forest in terms of stand height and density at each location. However, the challenge in regional-scale  138 O ’ NEAL   ET   AL .temperature modeling is that an actively logged terrain presents a patchwork land-scape of stands with different areal extents in various stages of growth, with eachstand having a unique surface energy balance and temperature. O’Neal et al. (2009) produced a model of regional near-surface temperature foran actively logged terrain in the Cascade Range of western Washington usingremotely sensed imagery, digital elevation model data, and field sensor data from aclearcut and an adjacent mature forest. That model relied on the following assump-tions: (1) a proxy record of stand height and density accurately represents the arrayof natural vegetation among the forest endmember selections; (2) there is a domi-nant regional climate pattern that affects all their study sites and their modeldomain; and (3) the average environmental lapse rate of 7° C km –1  is appropriate forthe Cascade Range study area.Our purpose in this study is to expand on O’Neal et al. (2009) by: (1) testingassumptions regarding the relationship between forest height and density and tem-perature used in previous studies; and (2) developing a refined regional temperaturemodel for a 15 × 24 km study area near Mount Adams, Washington, grounded innew field data. We first investigate the relationship between stand height and den-sity and temperature by collecting data from 13 different forest stands at differentelevations. Second, we assess the accuracy of the shade-based stand height anddensity record obtained from analysis of remotely sensed imagery by comparingthose results with field observations. Finally, we determine a field-based lapse ratethat can be combined with the aforementioned data to produce near-surfacetemperature models of the modern, actively logged study area and a modeled pre-anthropogenic forested landscape. We compare these models to estimate theeffects of logging on near-surface atmospheric temperatures.BACKGROUND Landcover Change in the Cascade Range of Western Washington Forestry practices in western Washington have significantly reduced the amountof mature and old-growth forests in the region. Strittholt et al. (2006) estimated that,before European settlement, approximately two-thirds of the total land area in theregion was old-growth forest. However, due to recent intensive forestry practices,approximately 70 to 90% of the srcinal conifer old-growth forests have been lost.At historic rates of reduction, most old-growth forests would have been eliminatedwithin the first few decades of the 21st century (Franklin et al., 1981). In 1994, theNorthwest Forest Plan (NWFP) was implemented for the Pacific Northwest becausedeclines in the amount of old-growth forests were exceptionally high in previousdecades. The plan provided management of late-successional and old-growth for-ests to preserve biodiversity. As a result, the annual increase of late-successionaland old-growth forest in 2004 was 1.9%, higher than expected (Thomas et al.,2006). Although the plan reduced the cutting of old-growth forests on federal landsby approximately 80%, historically and actively logged terrains often do not fullymature before harvesting and consist entirely of varying stages of regrowth(Wimberly et al., 2000).  EFFECTS   OF   LANDCOVER   CHANGES   ON   DAYTIME   SUMMER   TEMPERATURES 139 The Physics of a Changing Energy Balance as a Result of Forest Reduction The surface energy balance where natural vegetation is removed and replacedwith industrial materials (i.e., concrete, asphalt, steel, and glass) is well studied inurban areas (Oke, 1988; Grimmond and Oke, 1995; Du et al., 2007). In a mannersimilar to urbanization, removal of natural vegetation from forested landscapesexposes soil, rock, or dry vegetation and alters the diurnal change in the surfaceradiation and energy budget while also altering the coupling between the surfaceand the atmosphere for energy and moisture exchanges.Primary inputs to the surface are downward-directed shortwave and longwaveradiation. These are influenced by the sun and the atmosphere but not by surfaceconditions. Radiative outputs include reflected radiation, primarily consisting of albedo multiplying downward-directed shortwave, and longwave reradiation,which is mostly a function of surface temperature. Thus, surface conditions do notaffect the radiative inputs but strongly affect the radiative outflows. However, onaverage, radiative inputs to the surface are larger than the outputs, both globally(Kiehl and Trenberth, 1997) and locally within a forest (McCaughey, 1981), so addi-tional energy must be removed by nonradiative means.The nonradiative coupling between the surface and the atmosphere has fivecomponents: latent heat flux, sensible heat flux, ground heat flux, net latent andsensible energy storage within the biomass, and net chemical energy storage due tophotosynthesis. In forested regions, ground heat flux and the storage terms sum toapproximately 5 to 10% of the radiative excess, so latent and sensible heat fluxesare the terms with most of the energy available to balance other changes in surfaceconditions (McCaughey, 1985).Moisture availability accounts for much of the difference in nonradiative surfaceenergy between clearcut and mature-growth terrains. Hornbeck (1970) found thatin a West Virginia clearcut the amount of runoff increased, leaving less moistureavailable and thus less energy was used for evapotranspiration. Approximately one-fourth of the excess energy became upward thermal radiation, while the remainingthree-fourths were converted into sensible heat, increasing near-surface air temper-atures in the clearcut area. Although the clearcut will likely have a higher surfacealbedo, and hence less solar input, than the forest, a mature forest is an extremelyefficient mixing environment to transfer sensible and latent heat into the atmo-sphere (Klingaman et al., 2008). The daytime net radiation of a clearcut is signifi-cantly smaller than that of a forested site, while the nighttime net radiative loss of aclearcut is larger (Hornbeck, 1970; McCaughey, 1981). Despite the wide differ-ences in daytime energy balance conditions, the net radiation for a 24-hour periodmay be similar for both the clearcut and forested sites, with nighttime longwavereradiation balancing the system in both cases. The finding that clearcut areas should have higher afternoon temperatures undersunny conditions than nearby forested areas subject to the same regional weather,as described in the citations at the beginning of this paper, is thus supported by con-siderations of the energy balance components involved. Our purpose in the remain-der of the paper is to quantify the regional effect on temperature caused by theinfluence of clearcutting on the local energy balance.  140 O ’ NEAL   ET   AL .STUDY AREAOur study area is a 15 × 24 km area on the west and northwest flanks of MountAdams in the Gifford Pinchot National Forest, located in the Cascades mountainrange of southwestern Washington state (Fig. 1). This study area is located 55 kmeast of Mount Saint Helens and 73 km south of Mount Rainier; the latitude rangesfrom 46°11' N to 46°24' N and the longitude ranges from 121°32' W to 121°43' W.Elevation ranges from approximately 500 m to 2000 m. Field surveys of the standsused for this study area revealed that all are mixed conifer types with no singledominant overstory species. Common species found in the study domain includeDouglas fir ( Pseudotsuga menziesii  ), Pacific silver fir (  Abies amabilis ) , grand fir Fig. 1.  A. Cascade Mountain range of southwestern Washington state. B. The study area on thenorthwestern side of Mount Adams.  EFFECTS   OF   LANDCOVER   CHANGES   ON   DAYTIME   SUMMER   TEMPERATURES 141(  Abies grandis ), and western and mountain hemlock ( Tsuga heterophylla and Tsugamertensiana ; Little, 1980). The area is an actively logged terrain with intermixedstands of varying ages, the oldest of which are dominated by trees estimated to beover 400 years in age (USDA, 1999).FIELD METHODSOne of the primary goals of this study was to better understand temperaturedifferences among forests of different stand densities at different elevations. Toaccomplish this, we placed temperature sensors at 13 field sites, selected toaccount for six groups of different forest ages. Although there is no absolute rela-tionship between stand ages and densities, we identified prospective sites usingforest inventory data from the USDA (1999) in combination with visual assessmentof topographic maps and high-resolution aerial imagery from the 2006 NationalAgriculture Imagery Program (NAIP) dataset (USDA, 2006). Age groups weredefined by 25-year intervals ranging from 0 to 25 years up to 125 years and older.The different forest age groups were then stratified into three elevation ranges (400–800 m, 800–1200 m, and 1200–1600 m) with the intent of having both age andelevation diversity in site selection. Potential sites were selected within 1 km fromthe road network to provide reasonable access to the sensors, and a final restrictionon potential sites was attained by avoiding locations within 0.5 km of large riversand lakes to limit local moisture effects on the temperature data. Final selection of sites could only be made during the actual field reconnaissance, leading to the 13sites that were established (Fig. 2). At each site, temperature sensors (HOBO Pro v2 data loggers) were placed atapproximately 1 m above the surface and at least 50 m inside the forest/clearcutboundary to prevent complications by forest fringe effects with other local environ-ments. Sensors were placed in areas that received no direct sunlight (in the forest)or in radiation shields (in the clearcut or regrowth sites) to prevent direct heating of the sensor. Each sensor collected temperature data at 10-minute intervals for 60days from July 20, 2008, through September 17, 2008. Although this period trun-cates the summer record, it includes only snow-free summer days. Snow coveraffects surface albedo and radiation budgets far beyond the vegetation effects thatare being studied; avoiding such effects was essential. Stand Density Data Because of the patchwork of nonuniform stand heights and densities in the studyarea, a regional model of temperature based on interpolating between the 13 sen-sors to a spatial field would have limited value. Therefore, we combined our fielddata with analyses of remotely sensed imagery to derive a forest stand-height anddensity relationship that can be extrapolated to our entire study area. Fortunately,several studies have shown the utility of space-borne imagery in developing modelsof forest canopy cover and successional stages on a regional scale (Roy et al., 1996;Boyd et al., 2002; Sabol et al., 2002; Song et al. 2007). Thus, the temperaturemodels developed for this study rely on a stand height and density proxy record
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks