干旱区生态环境知识资源中心

Climate change has huge impact on socio-economic and natural systems of semi-arid areas. Vegetation dynamics plays a crucial role in natural resources and land use planning and regional policy decisions. Increasing remote sensing platforms like satellites, airborne surveys, unmanned aerial vehicles, etc. facilitate the collection of spatiotemporal earth observations. The main objective of this study is to improve processing capabilities of proximal wireless sensing systems for crop water management in semi-arid areas. Methodology is proposed comprising of estimation of phenological stages of major Land Use / Land Cover (LU/LC) (i.e. forest, scrub forest and agriculture) in semi-arid region using multi-sensor (Landsat 7 and 8) remote sensing time series observations. A study area from semi-arid region of central India is selected to study the vegetation growth stages. In this study freely available non-commercial satellite imagery store Google Earth Engine (GEE) is used to extract the time series of remote sensing observations over the selected area. The time series of vegetation indices derived from satellite based reflectance has been analyzed to estimate the phenological stages of forest, agriculture and scrub forest. Weather variables like rainfall were used to validate the estimated vegetation growth stages. The proposed methodology attempts towards real / near real time Internet of Things (IoT) based on board sensor data processing to provide information on region specific subsistence irrigation schedules which would result into crop yield improvement.