EnMAP-BMP
| Objective | Quantify bio-methane potential of agricultural areas using remote sensing |
| Target Variables | Bio-methane potential (BMP), Vegetation structual parameters (Biomass, LAI), Land use |
| Datengrundlage | HyMAP, APEX, simlated EnMAP data, SAR, ASD-FieldSpec, Laboratory analyses |
| Description | The project goals are implemented by an interdisciplinary consortium with complementary expertise in the field of remote sensing, agriculture and ecosystem modeling. The EnMAP-BMP project aims to develop a method for the remote sensing of the biomethane potential (BMP) of energy crops (including maize). The analytical determination of the BMP on a laboratory scale takes up to 30 days and is therefore time-consuming and expensive. An alternative is the use of Near Infrared Spectroscopy (NIRS), with which the BMP can also be determined. The attempt to transfer this approach to in-situ conditions using hyperspectral remote sensing systems is therefore an obvious choice. In addition to airborne hyperspectral data (HyMAP and APEX) and EnMap data simulated from them, a mobile NIRS station is also used. Current regression methods such as PLSR and SVR are used to derive the BMP from the data. Regional application also requires the development of a suitable classifier (one-class classifier) in order to reliably differentiate energy crops from other land use classes. The derivation of the regional BMP also requires information on the biomass. Therefore, concepts of sensor fusion (SAR + hyperspectral) are taken into account in order to obtain an improved assessment of plant structural parameters (e.g. LAI and biomass). Among other things, the potential of ensemble-based procedures and approaches of decision fusion will be examined. In connection with a dynamic plant growth model, which provides continuous yield estimates, an additional regional estimate of the bioenergy potential will be carried out. The figure illustrates the workflow and data generation, as well as the cooperation of the working groups within the individual work packages. |
| Partners | 1) Prof. Dr. Thomas Udelhoven (Kontaktperson) 2) Prof. Björn Waske (Universität Bonn) 3) Dr. Holger Lilienthal (Braunschweig) 5) Dr. Philippe Delfosse 4) Prof. Christoph Emmerling |
| Präsentations | Project presentation at 3rd National EnMAP-Workshop Präsentation of Bonn University at 3rd National EnMAP-Workshop |
| References | Delfosse, P., Lemaigre, S., Flammang, J., Neuberg, C., Hausman, J.F., and Hoffmann, L., (1010): Determination of the Biomethan Potential of plant residues using Near Infraread Spectroscopy, in preparation. Delfosse, P., Lemaigre, S., Flammang, J., Neuberg, C., Hausman, J.F., and Hoffmann, L., (2007): Evaluation variétale du maïs, du tournesol, et du sorgho pour la méthanisation au Grand-Duché de Luxembourg. IN: Biométhanisation Agricole, proceedings of a one day meeting at Redange, 13 september 2007, Centre de Recherche public – Gabriel Lippmann & Administration des Services Techniques de l’Agriculture, Grand-Duchy of Luxembourg. Udelhoven, T., van der Linden, S., Waske, B., Stellmes, M., and Hoffmann, L. (2009): Large area mapping using hypertemporal data and decision fusion. Geoscience and Remote Sensing Letters, 6, 592-596. Udelhoven, T., Atzberger, C. and Hill, J. (2000): Retrieving Structural and Biochemical Forest Characteristics Using Artificial Neural Networks and Physically Based Reflectance Models. in Buchroithner, M. (Ed.), EARSeL Symposium, 14-16 June 2000, Dresden, 205-212. Voivontas, D., Assimacopoulos, D., and Koukios, E. G. (2001): Assessment of biomass potential for power production: a GIS based method, Biomass and Bioenergy, 20, 101-112. Waske, B., Benediktsson, J.A. (2007): Fusion of Support Vector Machines for Classification of Mulitsensor Data. IEEE Trans. on Geoscience and Remote Sensing, 45, 3858-3866. |