B081-0014
Predicting beneath canopy radiation for land surface modelling from voxelised terrestrial laser scanning data

Monday, 14 December 2020
Poster
Steven Hancock1, Matthew Purslow1, Richard Essery2, Tobias Jonas3, Giulia Mazzotti4, Johanna Malle5, Clare Webster4, Caio Hamamura6, Rachel Gaulton7, Zemin Feng8, Zhanliang Zhu8, Jonathan A Greenberg9, Theodore Elliott Hartsook9, Katelyn Josifko9, Adriano Matos9 and Laura Wade9, (1)University of Edinburgh, Edinburgh, EH9, United Kingdom, (2)University of Edinburgh, School of GeoSciences, Edinburgh, United Kingdom, (3)SLF / WSL, Davos Dorf, Switzerland, (4)WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland, (5)Northumbria University, Newcastle-Upon-Tyne, United Kingdom, (6)Federal Institute of Education, Science and Technology of São Paulo, São Paulo, Brazil, (7)University of Newcastle, Newcastle, United Kingdom, (8)University of Edinburgh, Edinburgh, United Kingdom, (9)University of Nevada Reno, Reno, United States
Abstract:
Radiative transfer (RT) of sunlight within vegetation canopies controls albedo, snowmelt and photosynthesis. Vegetation causes shadowing, multiple scattering and absorption to create a heterogeneous light regime. Land surface model (LSM) RT schemes largely ignore this heterogeneity and use homogeneous RT models. In reality, heterogeneity causes the ground and leaves to be illuminated with a range of intensities and due to the non-linearity of processes, heterogeneous RT models predict different fluxes than homogeneous models.

A new generation of satellites can characterise global vegetation structure. In order to understand the link between structure and heterogeneous light environments at a global scale, a heterogeneous RT model is needed along with high-resolution measurements of structure to drive it. Such a model can also predict satellite data, allowing a heterogeneous RT model to be driven by globally available data.

This study makes use of an open-source library for estimating voxelised gap fraction and leaf area index (LAI) from terrestrial laser scanning (TLS) data, and then runs the same library backwards to produce an efficient heterogeneous RT model. The accuracy of the TLS estimate of voxelised LAI is validated against destructive harvesting data and the open-source RT model is validated against radiometer data. The impacts of voxel size, occlusion and multiple-scattering on predicted RT are explored. The new RT model will be benchmarked against established voxel RT models. The RT regime predicted by LAI is compared to that predicted by gap fraction voxels to see whether it is necessary to convert gap fraction (which is directly measured by TLS) to LAI (which is inverted via a model) for driving an RT model. Initial results suggest that each voxel needs to be at least 60% visible to a TLS scan location to allow an accurate gap fraction and LAI estimate, and the accuracy of spatially and temporally explicit beneath canopy radiation is quantified.

<< Previous Abstract | Next Abstract >>