Agricultural production is an undertaking that human beings create agricultural products by making use of organismic growth and development rules, a process tends to be restricted and influenced by the natural environment. To make better use of agricultural resources, investigation and monitoring of soil properties, geology, climate change, layout of crops planted and their growth and development in various ways are necessary.
Traditional monitoring of agricultural planting structures relies mainly on labor surveys which are comparatively disadvantageous in conduction frequency, human costs, timeliness and accuracy and no longer meets the demands of the modern agriculture.
A New Solution
Satellite RS technology featuring broad coverage, short revisiting period and cost-efficient acquisition plays an important role in investigation, evaluation, monitoring and management of large-scale outdoor agricultural production. It has already been used in a wide range of fields including crop identification, area reckoning, crop classification, yield estimation, growth and diseases & pests monitoring, etc.
The agricultural planting structure monitoring of the Nonbashi Shihezi Farm in northwest China’s Xinjiang Uygur Autonomous Region is taken as an example of the application of RS in agricultural and crops distribution investigation.
Data extraction is completed based on spectral, texture, shape and spatial relationship traits, after the combination of image features (tone or color, also spectral patterns), spatial features (shape, size, shadow, texture, graphics, location and layout) with non-RS data (topographic and thematic maps) and analyses and augmented treatment of the spectral curves of the ground truth at each band.
We worked out the following thematic map by extracting the spatial distribution data of staple crops on the basis of combined high & medium resolution RS images, manifold spectral traits, spatial heterogeneity information and climate characteristics.
Spatial Distribution of Main Crops in Nonbashi Shihezi Farm in northwest China’s Xinjiang Uygur Autonomous Region in May, 2017
（1）Cotton is the most widely planted with forest, grapes and wheat in the wake.
（2） The crops distribution conforms to the local climate conditions. For example, the Xinjiang cotton, an early to medium or even special-early maturing species, is photophilous and suitable to grow in the sunny Xinjiang Autonomous Region. That’s the reason that cotton is the most widely planted crop here.
（3） Interval distribution between wheat and grape can be seen here, for their distinct requirements of soil and water qualities as well as a blockade of water and nutrients competition.
（4） Forests and woods are scattered in various regions, mainly around water areas.
We Can Do More
RS monitors the planting area, growth, yield estimation, soil moisture, diseases and pests and other crops information.
（1）Planting area monitoring:different crop differs in color, texture and shape on the RS images, and the data of crop planting area can be extracted to obtain crop planting area and scale.
（2）Growth monitoring: referring to the macroscopic monitoring of seedling, growing and changing of crops, namely to monitor the growing conditions and trends by LAI (Leaf Area Index) and NDVI (Normalized Differential Vegetation Index).
（3）Yield estimation: RS imagery is also able to monitor and forecast crop yield based on its unique spectral reflectance, which can be used to reversely deduct the crop growth (such as LAI and biomass). The output information could be acquired finally through the correlation model of growth data and yield.
（4）Soil moisture monitoring:, the spectral features of the same soil may vary according to different soil moisture content. RS monitoring of the content is via visible lights at near-infrared, thermal infrared and microwave bands. A relationship model is established with soil moisture content parameters to carry out reverse deduction of the moisture content.
（5） Plant diseasesand pestsmonitoring and forecast: the vegetation’s response to diseases and insect pests, lack of fertilizer and other threats varies with threat types and degrees including biochemical (cellulose, blade, ) and biophysical (canopy structure, coverage, LAI, etc.) changes, along with subsequent changes in spectral pattern of plants absorptivity. Thus it is possible to detect the threats at an early stage. Periodic extraction of the affected area and spatial distribution can be made via RS imagery/
With a superiority of fast macroscopic data acquisition, the RS technique makes it possible to evaluate and monitor the volume, quality and spatial distribution of agricultural natural resources including arable lands, grasslands and waters and thereby provides scientific support for the exploitation & protection of agricultural resources, agri-planning, eco-environmental protection and sustainable development of agriculture.
The RS technique is an important technical means of disaster relief monitoring and evaluation. It is capable of dynamically monitoring major agricultural and natural disasters like droughts and floods, including their occurrence, impacted spheres, affected area & degree, contributing to the pre-disaster warning, post-disaster relief and alleviate the damages to the agricultural production.