PhD Defence by Mr Joris Timmermans
Dept. of Water Resources
Title of Defence
Coupling optical and thermal radiative transfer to biophysical processes in vegetated canopies
Summary
In this dissertation I have investigated the potential of using optical and thermal multi-directional remote sensing observations to estimate biophysical processes. Monitoring the processes within the earth’s biosphere is extremely important. Remote sensing is probably the only tool for monitoring such processes at regional to continental scale. Presently several initiatives exist to estimate evapotranspiration and photosynthesis using remote sensing algorithms. Most of these algorithms have not been developed recently, and may already exist for some years. This has the advantage that these algorithms have been tested and validated in several ways during the past couple of years. However, using older models also limits the potential of using new innovative approaches in satellite sensors.
In the last few years the quality, availability and innovations of remote sensing sensors have increased remarkably. Remote sensing products can now be produced with higher spatial, temporal, and spectral resolutions than previously thought possible. All these innovations offer the possibility to estimate biophysical processes at higher accuracies. In addition, some of these new satellite sensors even offer the option of acquiring observations from different viewing directions.
Even though new satellite sensors have an increased spatial resolution, pixel resolutions are still several orders above that of individual canopy components. The radiation observed by the satellite sensor is therefore an aggregate of the (reflected / thermal emitted) radiation by these components. This aggregate depends on the structure of the canopy, the spectral characteristics of the components, and the observation/solar angles. Consequently this aggregate will have an angular signature, such as the bidirectional reflectance factor, which may be measured. Knowledge about subpixel processes can therefore be obtained on the basis of these angular signatures. Radiative transfer models are required to characterise the behaviour of an angular signature on the basis of different canopy parameters. Because current remote sensing algorithms do not employ complex radiative transfer models, they are not yet suited to making use of multi-directional observations.
In this dissertation a model is presented, which is able to simulate the biophysical processes, the radiative transfer within the canopy and the outgoing hyperspectral optical and thermal radiances in parallel. This Soil Canopy Observation of Photochemistry and Energy fluxes (SCOPE) model combines a discrete version of the Scattering by Arbitrary Inclined Leaves (SAIL) 1D radiative transfer model, with a photosynthesis model, and an evaporation/transpiration model. The coupling of radiative transfer model with the biophysical process model is performed using the skin temperatures of four different canopy components, which are vertically distributed in the canopy. The fluxes are than calculated iteratively by solving an energy balance approach which integrated the emitted fluxes of sunlit and shaded leaves for all layers.

Identification of the components was based on the radiative and contact measurements of different vegetation types, i.e. short and tall grass, young and mature maize, wheat, barley and vineyard. Based on the directional radiative signatures the most complex vegetation type was found to be vineyard. In this crop the most important components were sunlit soil, shaded soil, sunlit leaves and shaded leaves. Vertical variation of these temperatures was of a lower order than the differences between the individual temperatures.
The radiative transfer part of the model was evaluated by comparing it to more complex radiative transfer models. Validation of the other sub models was not performed in this research. First a comparison was made between the Soil Leaf Canopy (SLC) 1D radiative transfer model and SCOPE. The SCOPE and SLC both have evolved from the SAIL radiative transfer model. However, SLC provides an analytical solution to the radiative transfer problem while SCOPE solves it numerically, in order to estimate leaf temperatures at different vertical positions within the canopy. The SLC model is also capable of simulating crown clumping effects within the canopy. The comparison between SCOPE and SLC showed that in SCOPE the finite size of the layers caused large errors for large oblique viewing angles. A modification was made to the SCOPE model, which solved this problem.
A second comparison was performed between SCOPE and the very complex Discrete Anisotropic Radiative Transfer (DART) 3D radiative transfer model. Several scenarios were simulated by the DART model with increasing structural complexity. The input data required by DART were obtained through a terrestrial laser scan of the Speulderbos forested area. To provide these input data for DART several 3D filter operations were performed on the point cloud data of this laser scan. The SCOPE DART comparison showed a good agreement between the two models for all scenarios but the ones with more extreme clumping. While for higher clumping (lower crown coverage) the SCOPE model failed to approximate the DART results the correctly, the SLC model was able to simulate bidirectional reflectance factors more in agreement with DART. Only when the crown cover was so low that the DART BRDF started to have solar azimuth angle dependencies, the SLC model could not reproduce this artefact of DART.
After the validation, the SCOPE model was used to estimate the skin temperatures of the different canopy components. For this a sensor simulator was created to convert SCOPE top-of-canopy radiances into sensor observations. A sensitivity analysis proved that single-view (nadir) observations could not be used to estimate the skin temperatures of the different components; instead multi-directional observations were required. Using these simulated multi-directional observations, the skin temperatures of all four components could be retrieved successfully. Application of this algorithm with the multidirectional radiative field measurements also provided good results.

Finally, SCOPE was used to evaluate the Surface Energy Balance System (SEBS) remote sensing algorithm over tall vegetation. SCOPE was used here both to provide reference estimations of the turbulent heat flux and to simulate sensor bands required by the SEBS preprocessor. Comparison showed that SEBS underestimated the evapotranspiration, because the parameterization of the roughness length for heat was not suitable for tall vegetation. A modification was incorporated in the SEBS model that solved this problem. In addition, the evaporative fractions of SEBS and SCOPE were examined. It was found that the evaporative fraction calculated by SCOPE increased over the course of a day. Consequently the values of the evaporative fraction estimated using remote sensing observations can vary depending on the overpass time. Fortunately, the largest variation occurs very early in the morning. It was found that the mean evaporative fraction estimated by SCOPE matched the evaporative fraction estimated by SEBS at overpass time (between 9 h00 and 12h00).
It was a good run!
Biography
Joris Timmermans was as a twin born on 12 January 1979 in Nijmegen, Gelderland, Netherlands. He achieved his high school degree (VWO) in 1997 at the Canisius College, in Nijmegen (the Netherlands). Pursuing an academic career in Applied Physics at to the University of Twente, he moved to Enschede. His studies were focused on Physics of Fluids with small minors on Non-linear Optics and High Energy Physics. During this study he performed an internship in 2003 at the von Karman Institute in Brussels (Belgium) on the effect of instabilities in the Ariane booster rockets of the European Space Agency (ESA), he achieved his Master of Science (M.Sc.) degree in 2004 on the topic of ‘effects of ultrasound on micro bubbles behavior’. Following a life long ambition to do space research, he joined the International Institute for Geoinformation sciences and Earth Observation (ITC) in 2005 to perform his doctoral studies. During his research he performed several field experiments in the Netherlands, Germany and Spain with emphasis on Directional Radiative Acquisitions using a Goniometer. These field experiments have lead to the joint-development of the SCOPE (Soil Canopy Observation of Photochemistry and Energy Fluxes) ecological radiative transfer model. This model is currently used by him and several affiliated institutes and universities for simulating of biophysical and radiative transfer processes and for validation of remote sensing algorithms. He currently is finishing the work for the ESA`s WACMOS project, in which he created a high resolution global evaporation product. He will continue to perform research at the ITC on the combination of radiative transfer, biophysical processes and high performance computing.
Timmermans, J., Su, Z. (Promotor) , Verhoef, W. (Promotor) and van der Tol, C. (assistant promotor) (2011) Coupling optical and thermal directional radiative transfer to biophysical processes in vegetated canopies : e-book. PhD thesis University of Twente; summaries in Dutch and English. ITC Dissertation 193, ISBN: 978-90-6164-313-5
| Timesheet | |
|---|---|
| Event starts: | Wednesday 06 July 2011 at 14:30 |
| Venue: | UT, Waaier 4 |
| City where event takes place: | Enschede |
| Country where event takes place: | Netherlands |