DRAGON4

Land degradation is a process by which the land productive capacity declines or even is completely lost under the influence of natural forces and human activities. The scope of land degradation has become global in the last decades, which compromises sustainable land management and threatens the safety of food production. Assessment and monitoring of land degradation based on Earth Observation make the most accepted scientific basis for controlling land degradation. Objectives:(1)To develop methods of vegetation & soil bio-physical variables(PV/NPV fractional cover, NPP, Soil organic matter and soil texture) retrieval based on satellite data in drylands, at local and regional scale;(2)To enhance, benchmark and validate two novel approaches to land degradation surveillance by remote sensing: a two-dimensional implementation of Rain Use Efficiency (2dRUE), and a Moisture-responded Net Primary Productivity (MNPP).(3)To use the said approaches to map land degradation in a study area defined by the Potential Extent of Desertification in China(PEDC). This is a delimitation of UNCCD-affected areas in terms of drylands within China.Methods:(1)Vegetation PV/NPV fractional cover, NPP estimation in drylands: Aiming to retrieve the PV/NPV fractional cover in arid regions, the non-linear mixture method will be applied in order to improve the spectral mixture models for PV/NPV fractional cover estimation based on field measured spectra and hyperspectral satellite data Proba-1. High spatial resolution (GF-1 and UAV) and field survey data would be used as reference data for validation. For NPP estimation, some key parameters in CASA model, including the FPAR and the light use efficiency, will be optimized to promote the accuracy of NPP estimation.(2)Soil organic matter and texture(sand, silt and clay) estimation in drylands: Parametric or nonparametric regression models will be studied to establish a good relationship between soil proprieties and soil spectra, and the model would be applied to PROBA-1, Sentinel-2 and GF-1 data. The influence of spectral, spatial resolution and the presence of vegetation cover on soil proprieties retrieval would be analyzed.(3) 2dRUE for land degradation assessment: Input data for this method are time series of a remotely sensed vegetation density parameter (typically, NDVI or NPP), corresponding climate fields, and a land use/land cover map. Both assessment of land states of ecological maturity, and monitoring of land productivity trends, are performed on this data set for the study period. The former are computed after a graphical representation of RUE against aridity index. The latter are derived from a stepwise regression analysis. (4) MNPP method for land degradation assessment: A MNPPmethod for identifying areas of land degradation based on the change of annual NPP and MNPP over time and MI will be developed. Method of severity assessment of Land degradation will be studied on the basis of modelling of relative benchmarks of pixel-based land degradation. Deliverables: The deliverables will be scientific outputs in terms of peer reviewed journal publications, PhD and MSc theses and data sets in terms of datasets, map and assessment system for land degradation. Availability of funding to run the project: During the 4 years research, several funding projects are being undertaken by the team members. They will support the joint researches of proposed the Dragon 4 cooperation project and the progress reporting in the annual symposium