Policy Update | Using Hyperspectral Remote Sensing Technology to Support the Planning of a Strong Agricultural Nation

The core policy direction and the precise focus of CAS Xiguang Aerospace

according to Plan for Accelerating the Construction of a Strong Agricultural Nation (2024-2035)The requirement is that the construction of a strong agricultural nation will focus on Smart agriculture, efficient resource utilization, disaster prevention and control, and food securityFour core areas. Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD relies on... Hyperspectral remote sensing, satellite big data and AI algorithmsEmpowering you comprehensively from the following dimensions:

Smart Agriculture: Technology Empowers Precision Management

Policy requirements: Accelerate the construction of an integrated sky-ground agricultural observation network to promote the digitalization and intelligentization of agricultural production.

Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD Solution:

Hyperspectral remote sensing monitoring: Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD Relying on independently developed hyperspectral satellite constellations, UAV remote sensing clusters, and ground-based Internet of Things sensor networks, an integrated "sky-ground" agricultural observation system is being constructed.,pass Multi-source data fusion (0.4-2.5μm hyperspectral imagery + real-time meteorological/soil sensing), combined with AI algorithms (dual-driven deep learning and physical models).,accomplish Crop growth dynamic monitoring 、 Diagnosis of nitrogen, phosphorus and potassium nutritional deficiencies 、 Early identification of diseases and pests using the spectrumAnd in conjunction with the agricultural big data platform, generate " "One Field, One Policy" Health Diagnosis Report,support Variable irrigationand Disaster early warningdecision making.

Efficient use of resources: Safeguarding the red line of arable land and ecological security

Policy requirements: Strengthen the monitoring of arable land quality and promote the economical and intensive use of water and soil resources. Improve the mechanism for accepting and verifying the quality of supplementary arable land, and firmly safeguard the red line for the protection of 1.865 billion mu of arable land and 1.546 billion mu of permanent basic farmland.

Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD Solution:

1) Weed and soil moisture monitoring: Accurate identification of soil, crop, and weed distribution based on hyperspectral technology Real-time analysis of soil moisture 、 Moisture contentand Vegetation Index (NDVI) ， Dynamically constructing a coupled model of soil moisture and weeds in farmland,for Precision restoration of degraded farmlandand Water-saving irrigationProvide data support.

2) Ecological red line supervision: Based on multi-source satellite remote sensing technology, a high-frequency dynamic monitoring network is constructed. Through high-resolution image segmentation and texture feature analysis, the spatial boundaries of illegally occupied farmland and forest land and traces of surface changes are accurately identified.

By combining deep learning algorithms, complex terrain and irregularly shaped vegetation plots are adaptively extracted to generate illegal plots and law enforcement early warning reports, providing natural resources departments with all-weather, large-scale "sky eye" monitoring capabilities and supporting a closed-loop law enforcement mechanism of "early detection and rapid disposal".

Object-oriented classification is achieved based on multi-scale segmentation technology, spectral-texture features are integrated to optimize land cover extraction, high reflectivity targets and sandy land are merged to improve classification accuracy, and a high-confidence land cover distribution map is generated. A spatiotemporal evolution model is constructed by overlaying multi-period historical remote sensing data to dynamically track urban expansion and land use transformation trends.

Disaster Prevention and Control: Building a Strong Risk Barrier for Agriculture

Policy requirements:Establish a sound monitoring and early warning system for the entire supply chain of grain and important agricultural products. Improve the agricultural disaster early warning system and enhance disaster prevention and mitigation capabilities.

Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD Solution:

1) Identification of areas with abnormal growth:Based on the fusion of visible light and multispectral data from UAVs, and through orthorectification, vegetation index inversion and unsupervised clustering, a crop growth heat map (with red and yellow markings indicating good and bad areas) is generated, enabling accurate identification and visual monitoring of abnormal growth areas across the entire region.

2) Drought and flood disaster early warning:By combining multi-source remote sensing data fusion (optical/thermal infrared/microwave), a drought and flood disaster early warning model is constructed:

Soil moisture inversion:An inversion model is established using soil moisture-sensitive spectral bands (SWIR/NIR) to generate a high-precision soil moisture distribution map and analyze the regional water deficit or saturation state.

Drought dynamic monitoring:A drought index (such as TVDI) is constructed by combining land surface temperature (LST) and vegetation index (NDVI) to quantify the drought level (LST↑+NDVI↓ indicates severe drought).

Ecological index coupling:By integrating multiple parameters such as vegetation greenness, humidity component, and thermal environment, a remote sensing ecological index (RSEI) is constructed to assess farmland ecological pressure in real time and identify high-risk areas for drought and flood.

3) Early identification of diseases and pests:Based on hyperspectral remote sensing technology, early signals of pests and diseases in the egg stage can be identified by the variation characteristics of leaf spectral reflectance (accuracy > 85%), accurately locating outbreak areas and improving control response efficiency by 50%. Combined with multi-temporal remote sensing imagery, the spatial distribution, damage level, and spread paths of pests and diseases can be monitored in real time. A population dynamics model can be constructed to predict the spread trend of disasters, providing decision support for precision pesticide application and regional joint prevention and control, helping to reduce control costs by more than 30%, and promoting agricultural disaster prevention from passive response to proactive control.

Food security: Science and technology safeguard China's food supply.

Policy requirements:We will make solid progress in the new round of action to increase grain production capacity by 100 billion jin and accelerate the construction of a national food security industrial belt.

Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD Solution:

1) Production forecasting model:Based on the fusion of multi-source remote sensing vegetation indices and historical data, a staple food yield prediction model is constructed to predict the total yield 60 days in advance and analyze the spatial heterogeneity of the entire region (location of high-yield/low-yield patches); coupled with meteorological and soil parameters, a yield potential classification map is generated to support precision agricultural management and optimal resource allocation.

2) Crop maturity monitoring:Based on the inversion of multi-dimensional physiological parameters (such as anthocyanin and carotenoid content) using hyperspectral remote sensing, the spatial heterogeneity of crop maturity is accurately analyzed; coupled with meteorological data to dynamically monitor grain moisture content, a maturity-moisture synergistic model is constructed to predict the optimal harvesting window for each region, shortening the operation cycle by 15% and avoiding precipitation risks, thereby achieving a double improvement in harvesting efficiency and yield.

3) Construction of high-standard farmland:based on 3D remote sensing modelingaccomplish High-precision digital twins of farmland enable intelligent planning of ditch and road network layout.Supporting the digital management of "tens of thousands of acres of fertile land" Increase land utilization by 20% and increase production capacity by 15%.

CAS Xi'an Aerospace: Resonating with National Strategy

To date, CAS Xiguang Aerospace has served Five provinces across the country,cover More than 500,000 mu of arable landThis has cumulatively reduced costs and increased efficiency for agricultural entities. Over 75 million yuanIn the future, we will continue to innovate technology application scenarios and fully support the rapid implementation of the goals of the "Plan for Accelerating the Construction of a Strong Agricultural Nation (2024-2035)".