Local-scale heterogeneity of photosynthetically active radiation (PAR), absorbed PAR and net radiation as a function of topography, sky conditions and leaf area index

https://doi.org/10.1016/j.rse.2005.09.021Get rights and content

Abstract

The local-scale spatial distribution of photosynthetically active radiation (PAR), absorbed PAR (APAR) and net all-wave radiation (Q) across the top of a forest canopy was investigated as a function of topography, sky conditions and forest heterogeneity for a forested hilly study site located in south-central Indiana, USA that is part of the FLUXNET and SpecNet networks. The method to estimate spatial variability of radiation components utilized theoretical radiation modeling applied to a topographic model combined with spatial distribution of leaf area index derived from IKONOS imagery and empirical models derived from data collected on a single flux tower. Modeled PAR and Q compared consistently well with observations from a single tower with differences typically less than 10%, although clear-sky conditions were simulated more accurately than cloudy conditions. Spatial variability of radiation was found to be very sensitive to topographic relief and could be scaled linearly by mean slope angle. Decreases in optical transmissivity and increases in cloudiness had a strong effect of reducing both the spatial average and standard deviation of radiation components. Spatial variability of APAR was 53% greater than PAR and the characteristic scale of variance was reduced due to finer scale and magnitude of variance of LAI. Clear seasonal patterns existed in both spatial average and standard deviation values with summer producing the largest mean values and weakest spatial variability due to smaller solar zenith angles and seasonality in both optical transmissivity (scaled linearly by specific humidity) and cloudiness. These findings of spatial variability illustrate the need to characterize the complex landscape patterns at flux tower sites, particularly where the goal is to relate flux tower data to satellite imagery.

Introduction

Radiation fluxes at the Earth's surface play important roles in many ecological, climatological, and hydrological systems. Biological activity is strongly dependent on radiative transfer both directly, through the interaction between phytoelements and radiant energy emitted by the sun, and indirectly through micrometeorological controls. Understanding the spatial distribution of photosynthetically active radiation (PAR) is important for predicting patterns of ecosystem functioning within a forest (Vierling & Wessman, 2000) and gross ecological production (GEP) has been directly linked with PAR in numerous studies (e.g. Goulden et al., 1997, Gu et al., 1999, Monteith, 1972, Oliphant et al., 2002, Schmid et al., 2000). Improvements in remote sensing technology and techniques to estimate photosynthetic activity using hyperspectral sensors can be used to assess small scale spatial variability of ecosystem dynamics (Blackburn, 1999, Gamon et al., 1993, Rahman et al., 2001, Rahman et al., 2003). Net all-wave radiation provides the fundamental input to the surface energy balance and spatial variability has been linked with a variety of biometeorological controls including surface climates (Kalthoff et al., 1999) and rates of evapotranspiration (Famiglietti & Wood, 1995) as well as boundary layer processes and local thermal circulations (Kossmann et al., 2002). Since seasonality plays a critical role in temporal variability of ecosystem functioning, it is also important to observe the role of seasons on spatial variability of radiation due to changing geometric relations, optical transmissivity and cloud conditions.

Current theoretical models and satellite derived data can provide valuable insight into spatial variability of surface radiation fluxes (Diak et al., 2004, Dubayah et al., 1990, Dubayah and Loechel, 1997, McKenny et al., 1999, Oliphant et al., 2003). Here we explore the use of theoretical models of surface radiation fluxes in complex topography, IKONOS imagery (visible and near-infrared) and empirical relations determined from data collected from a single flux tower to estimate components of spatial variability of (PAR), absorbed PAR (APAR) and net all-wave radiation (Q) for a forested hilly region of mid-western USA. The primary objective is to estimate local-scale spatial heterogeneity of radiative fluxes at the top of the forest canopy across a topographically complex area. In particular we examine the role of topography, forest heterogeneity, optical transmissivity, cloud cover and seasonality on spatial variability of radiation.

Combining ecosystem modeling, satellite image interpretation and point observations has greatly enhanced the spatial understanding of ecosystem–atmosphere interaction at the continental to global scale (Running et al., 1999), although there remains a need to improve understanding of small-scale variability to inform downscaling of regional or larger models. Furthermore, characterization of ecosystem–atmosphere interaction is often based on data collected at individual towers using the eddy covariance approach (e.g. Schmid et al., 2000) as well as optical sampling techniques (Gamon et al. this issue). Assessing the spatial distribution of radiation components is therefore an important step toward scaling up of point measurements to regional-scale ecosystem estimates and variability within the source areas of turbulent flux measurements.

Section snippets

Site description and observational data

The location for this research was the Morgan–Monroe State Forest (MMSF) in south-central Indiana, mid-western USA (39°19′N, 86°25′W). MMSF is an extensive managed forest with a total area of 95.3 km2 (Schmid et al., 2000). The area used for model and image analysis in this study is a 3.5 × 3.5 km area of near-contiguous forest cover. At the center of this area on a ridge top with unobstructed sky view, is a 46-m instrumented AmeriFlux tower. Fig. 1 shows the shaded relief within the study area

Spatial model description

Surface radiation modeling in complex terrain has received growing attention in recent decades, although many studies have tended to focus on individual components, primarily incident shortwave radiation (Dubayah et al., 1990, Dubayah and Rich, 1995, Dubayah and van Katwijk, 1992, Kumar et al., 1997, Olseth and Skartveit, 2001, Whiteman, 1990, Whiteman and Allwine, 1986) or net shortwave radiation (Dubayah & Loechel, 1997), but including longwave radiation (Marks & Dozier, 1979), and PAR (

Results

This section reports on spatial variability of radiation fluxes found as a function of topography, cloudiness, seasonality and surface heterogeneity. For this analysis, simulations included empirical inputs from data collected from 1998 to 2001.

Conclusions

Potential spatial variability of PAR and Q were investigated for a hilly mid-latitude forested region in southern Indiana, in order to identify the relative controls of local scale topography, forest heterogeneity and atmospheric conditions. Simulations were run for observed cloudiness as well as the equivalent clear-sky conditions. Monthly mean model output compared well with observed point data with differences typically less than 10% and clear-sky days were simulated more accurately than

Acknowledgements

Funding for this research was provided by NIGEC, Dept. of Energy (Co-operative Agreement No. DE-FC03-90ER61010). The authors would like to thank the large number of people involved in data collection at the MMSF tower site, particularly Steve Scott and Brian Offerle.

References (53)

  • S.W. Running et al.

    A global terrestrial monitoring network integrating tower fluxes, flask sampling, ecosystem modeling and EOS satellite data

    Remote Sensing of Environment

    (1999, Oct)
  • H.P. Schmid et al.

    Measurements of CO2 and energy fluxes over a mixed hardwood forest in the mid-western United States

    Agricultural and Forest Meteorology

    (2000)
  • S. Wang et al.

    New calculation methods of diurnal distribution of solar radiation and its interception by canopy over complex terrain

    Ecological Modelling

    (2002)
  • C.D. Whiteman et al.

    Extraterrestrial solar radiation on inclined surfaces

    Environmental Software

    (1986)
  • M.A. Wulder et al.

    Aerial image texture information in the estimation of northern deciduous and mixed wood forest leaf area index (LAI)

    Remote Sensing of Environment

    (1998)
  • L.A. Vierling et al.

    Photosynthetically active radiation heterogeneity within a monodominant Congolese rain forest canopy

    Agricultural and Forest Meteorology

    (2000)
  • I. Alados et al.

    Direct and diffuse photosynthetically active radiation: Measurements and modeling

    Agricultural and Forest Meteorology

    (1999)
  • G.R. Diak et al.

    Estimating land surface energy budgets from space: Review and Current Efforts at the University of Wisconsin–Madison and USDA–ARS

    Bulletin of the American Meteorological Society

    (2004)
  • Dubayah, R. (1991). The topographic variability of solar radiation. Unpublished PhD dissertation, Department of...
  • R. Dubayah et al.

    Topographic distribution of clear-sky radiation over the Konza Praire, Kansas, USA

    Water Resources Research

    (1990)
  • R. Dubayah et al.

    Modeling topographic solar radiation using GOES data

    Journal of Applied Meteorology

    (1997)
  • R. Dubayah et al.

    Topographic solar radiation models for GIS

    International Journal of Geographical Information Systems

    (1995)
  • R. Dubayah et al.

    The topographic distribution of annual incoming solar radiation in the Rio Grande River basin

    Geophysical Research Letters

    (1992)
  • J.L. Ehman et al.

    An initial intercomparison of micrometeorological and ecological inventory estimates of carbon exchange in a mid-latitude deciduous forest

    Global Change Biology

    (2002)
  • J.S. Famiglietti et al.

    Effects of spatial variability and scale on areally averaged evapotranspiration

    Water Resources Research

    (1995)
  • J.C. Gallant

    Modeling solar radiation in forests of southeastern Australia

  • Cited by (23)

    • Evaluating biochemical and biophysical characteristics of tropical deciduous forests of eastern India using remote sensing and in-situ parameter estimation

      2023, Remote Sensing Applications: Society and Environment
      Citation Excerpt :

      Because of its high sensitivity in plants, FAPAR is used as one of the major parameters in remote sensing estimates including photosynthetic activities (Mahadevan et al., 2008; Zhang et al., 2020). Some studies reported that background reflectance and vegetation growth period influence the relationship between FAPAR and GPP (Oliphant et al., 2006). LAI is influenced by processes such as evaporation, rainfall interception, transpiration, photosynthesis, and carbon flux.

    • Drivers of shortwave radiation fluxes in Arctic tundra across scales

      2017, Remote Sensing of Environment
      Citation Excerpt :

      As mosses may account for a significant part of tundra absorbed radiation and gross ecosystem production (Huemmrich et al., 2010), further research is required on how to accurately represent mosses in radiation transfer and land surface models. In agreement with Oliphant et al. (2006), we found that at the5 cm pixel and the patch scale albedo, transmittance, and fAPAR can be substantially different from mean values of the respective vegetation types (Fig. 5). We found that the scale of the variability in radiation budget components is much smaller than what we can observe from satellites, posing challenges for driving models from conventional remote sensing.

    • Spatial variation of canopy PRI with shadow fraction caused by leaf-level irradiation conditions

      2016, Remote Sensing of Environment
      Citation Excerpt :

      LUE is defined a posteriori over a certain time interval as the ratio of gross primary production (GPP) to photosynthetically active radiation absorbed by green canopy elements (APAR) (Gitelson & Gamon, 2015; Monteith, 1977) and is widely utilized in modeling contexts. Both PAR and APAR can vary significantly within a forest canopy and are proportional to, for example, canopy structural parameters (e.g., the distribution of leaf area density) and the angle of illumination (e.g., Gamon et al., 2001; Hall, Hilker & Coops, 2012; Oliphant, Susan, Grimmond, Schmid, & Wayson, 2006). Gamon et al. (2001) demonstrated that APAR alone can explain a significant part of the photosynthetic variation within plant canopies.

    • Spectral Network (SpecNet)-What is it and why do we need it?

      2006, Remote Sensing of Environment
      Citation Excerpt :

      The issue of spatial variability in surface properties across landscapes, and within flux tower footprints in particular, is a key issue being explored by the SpecNet community, particularly since flux measurements alone cannot resolve these spatial patterns. Oliphant et al. (2006-this issue) used IKONOS images in a detailed exploration of the spatial patterns of absorbed photosynthetically active radiation (APAR, a critical term within ecosystem carbon uptake models) for a deciduous forest within the vicinity of a flux tower. They found considerable variation in APAR due to cover, season, and sky conditions (diffuse vs. direct radiation) and discuss the implications of this variability for carbon flux.

    View all citing articles on Scopus
    View full text