Working Towards a Global-scale Vegetation Water Product from SMOS Optical Depth
Grant, Jennifer1; Scholze, Marko1; Williams, Mathew2; Wigneron, Jean-Pierre3; Kerr, Yann4
1Lund University, SWEDEN; 2University of Edinburgh, UNITED KINGDOM; 3INRA, FRANCE; 4CESBIO, FRANCE

Background

Currently, global-scale monitoring of vegetation properties by remote sensing is almost always based on the use of optical or infrared sensors. The information provided by these sensors is most often related to chlorophyll content, photosynthetic properties, spectral characteristics or amount and classification of biomass and land cover type. However, passive microwave remote sensing has the potential to offer unique and complementary information, namely on the water status of the terrestrial vegetation. This information can be derived from the passive microwave vegetation optical depth, which is related to vegetation water content, structure and biomass (Kirdyashev et al., 1979). The relationship with vegetation water is of strong interest, as this variable is one of the key controls on photosynthesis and transpiration. Vegetation water status will determine both carbon uptake and water use of the plant, through the common pathway of the two fluxes - the leaf stomata. In this way, vegetation water status forms a crucial link between the carbon and water cycles. Besides having clear applications in the field of agriculture, as it is directly related to plant water stress, it is also of strong interest for e.g. terrestrial biosphere and climate modelling. In this study, gravimetric vegetation water content was obtained from optical depth measurements made by ESA's Soil Moisture and Ocean Salinity (SMOS) mission. This was done for a number of global vegetation classes, with a view towards developing a temporally dynamic, global-scale vegetation water product in the near future. Such a product does currently not exist, and would offer important complementary information to well-known vegetation indices from optical remote sensing such as e.g. NDVI and LAI.

Methodology
The gravimetric vegetation water content (Mg) was obtained from SMOS optical depth (:ó) measurements by combining the effective medium approach (Wegmïller et al., 1994) with the vegetation dielectric constant model of Ulaby & El-Rayes (1987). As Mg gives the amount of water available per unit of fresh biomass, i.e. it is expressed in [kg

  • kg-1], it is strongly linked to vegetation water status. Preliminary results using this approach for a coniferous forest site in the US were previously presented by Grant et al. (2012). In a first step, this approach was used for each of the 11 different vegetation classes in the University of Maryland (UMD) global landcover classification scheme. Two FLUXNET sites representative of each vegetation class were selected, for calibration and validation respectively. Modelled values of leaf water potential (: x leaf) were used as a validation proxy for Mg, as neither Mg nor :xleaf is usually measured directly in (long-term) field experiments. :xleaf was previously found to be correlated with :o -derived M g by Wigneron (1993). It was modelled using the Soil-Plant-Atmosphere (SPA) model (Williams et al., 1996), which is the only model of its kind to explicitly model plant water transport. The optical depth values were obtained from SMOS Level 2 data, selecting data from the nearest most representative SMOS gridcell for each site. The flags provided in the Level 2 products were used for filtering and quality checks. Temperature and soil information were obtained from the FLUXNET in situ measurements. The vegetation indices LAI, NDVI and/or EVI, used as biomass proxies, were obtained from MODIS data. Further vegetation type-specific inputs necessary for the various models were obtained from the literature.

    Results

    The correlation found between :xleaf and Mg for each vegetation type was used to determine the best-fitting value of the structure parameter (Ap) for that vegetation type. Once the value of Ap was thus calibrated, the modelling approach was validated over a different site with the same vegetation type. The results of both calibration and validation are presented and discussed here, for each UMD global vegetation class.

    Outlook

    With the above approach being established, calibrated and validated, the way is clear for producing a new SMOS Level 4 user product, namely global-scale maps of vegetation water. This will be done in a next step, this time using the global-scale (0.25i ãgrid) SMOS Level 3 optical depth data as input. Consistent with the temporal resolution of the SMOS Level 3 products, the vegetation water maps can be produced on daily timescales and deliver morning (ascending orbits) and evening (descending orbits) values. This will be a new and unique vegetation product with a broad range of applications, as discussed in the introduction. It is directly related to multiple Land Challenges of ESA's Living Planet Programme, as well as to several ongoing ESA science studies. Finally, there are clear links to the scientific objectives of several ESA land missions, as well as to NASA's upcoming Soil Moisture Active Passive (SMAP) mission. Acknowledgements This work was co-funded by ESA's STSE Changing Earth Science Network as part of the VEGWAC project.

    References

    Grant, J.P., Wigneron, J.-P., Drusch, M., Williams, M., Law, B.E., Novello, N. and Kerr, Y. ''Investigating temporal variations in vegetation water content derived from SMOS optical depth,'' Proceedings of the IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2012. Kirdyashev, K.P., Chukhlantsev, A.A. and Shutko, A.M. ''Microwave radiation of the earth's surface in the presence of vegetation cover,'' Radiotechnika, (24), 256-264, 1979.
  • Ulaby, F.T. and El-Rayes, M.A. i°Microwave Dielectric Spectrum of Vegetation "C Part II: Dual-Dispersion Model,i± IEEE Transactions on Geoscience and Remote Sensing, (5), 550-557, 1987.
  • Wegmüller, U., Mätzler, C. and Njoku, E.G. i°Canopy Opacity Models,i ± Passive Microwave Remote Sensing of Land-Atmosphere Interactions, pp. 375-387, 1995.
  • Wigneron, J.-P. ''Modalisation de l'émission micro-onde d'un couvert végétal "C Mise en relation de la mesure hyperfréquence passive avec les échanges énergétiques et hydriques d'une culture de soja,'' PhD thesis, Université de Paul Sabatier de Toulouse, ch. 8, 1993.
  • Williams, M., Rastetter, E.B., Fernandes, D.N., Goulden, M.L., Wofsy, S.C., Shaver, G.R., Melillo, J.M., Munger, J.W., Fan, S.-M. and Nadelhoffer, K.J. ''Modelling the soil-plant-atmosphere continuum in a Quercus-Acer stand at Harvard Forest: the regulation of stomatal conductance by light, nitrogen and soil/plant hydraulic properties,'' Plant, Cell and Environment, (19), 911-927, 1996.