crop yield estimation by remote sensing
nongzuovvu Yao gan guehan (erop yield estimation by remot. Sens- ing) the process of estimating crop yield by using remote sensing information and remote sensing methods. Telesensing information refers to the instantaneous record of reflection and radiation information of crops and their environmental background obtained by using various sensors on various remote sensing platforms. Through computer processing, identification, classification, information extraction and other remote sensing methods, combined with mathematical statistics analysis and geoscience analysis, the final yield of crops is estimated. According to the different sources of remote sensing data, the yield estimation of agricultural crops can be divided into space remote sensing crop yield estimation and ground remote sensing crop yield estimation. The former includes crop yield estimation based on satellite data and crop yield estimation based on aerial survey data, which has a wide range and strong macro-view. The latter is based on the spectral information of crops obtained from the ground remote sensing platform, and the yield estimation range is small. Crop yield estimation by remote sensing includes dynamic monitoring of crop growth process, estimation of planting area, yield estimation per unit area and total output estimation. Dynamic monitoring of crop growth process is one of the important bases for yield estimation by remote sensing in space. Polar-orbiting meteorological satellites (TIROS-N series by NOAA in the United States, FY-1 in China, etc.) are the main tools for crop growth monitoring because of their short repeated scanning period and economy. Growth monitoring is realized by analyzing the change of vegetation index of remote sensing spectrum with time. The earliest and most widely used remote sensing information to measure the crop planting area is the multispectral scanner (MSS) data of American Landsat (1 ugly ndsat), and now the thematic imaging scanner (rrM) data with higher geometric resolution and the data of SRyT are mostly used. According to the spectral characteristics of crops in different growth periods and the agricultural calendar, the interpretation marks are established, and then the multi-spectral data are identified and classified by visual inspection and computer combination, supplemented and corrected by the measured data on the ground, and finally the planting area is calculated. In recent years, it is studied to estimate the planting area by combining NOAA's improved very high resolution radiometer (AVHRR) data with Landsat thematic imaging scanner (TM). The yield forecast is based on the analysis of the relationship between crop yield and various influencing factors and the establishment of a regression model. In the early days, the weather data of meteorological satellites were used as the main input of crop yield estimation model. After the early 198s, crop information was gradually extracted from remote sensing data, and a remote sensing yield estimation model or a remote sensing parameter model was established on the basis of analyzing the relationship between remote sensing spectral vegetation index and crop yield or agronomic parameters (such as leaf area coefficient). In order to improve the accuracy of yield forecast, the method of integrating the forecast results of various yield estimation models into the final yield is also adopted. The total yield can be obtained by multiplying the yield by the planting area, or by establishing a remote sensing yield estimation model based on the analysis of the relationship between the total yield and the total spectral index value. Crop yield estimation based on ground remote sensing is completed by measuring the field spectra of crops in different growth periods and establishing a regression model between spectral data and crop yield. In the mid-197s, LACIE in the United States and ARS (pure state-focused ARS) followed it, which set a precedent for crop yield estimation by remote sensing. Many countries in the world have generally carried out crop yield estimation by remote sensing. Crop yield estimation by remote sensing in China began in the early so's, mainly using American land satellites to carry out small-scale research; Since the mid-so's, the China Meteorological Bureau has taken the lead in carrying out large-scale winter wheat remote sensing monitoring and yield estimation research experiments in 11 provinces (autonomous regions and municipalities directly under the Central Government) using polar-orbiting meteorological satellites, and it was transferred to the Meteorological Bureau in 19 years. The system of the Academy of Sciences, agricultural departments and colleges and universities have also carried out research and experiments on remote sensing yield estimation of various crops using various remote sensing data. In the early oo era, the yield estimation operation system of major crops in key grain-producing areas has been built. The yield estimation accuracy of wheat is over 95%, and that of corn and rice is over 85%. Crop yield estimation by remote sensing has the characteristics of rapidity, macro-economy, objectivity and so on, and it can dynamically monitor the growth process of crops, eliminating the limitations of human interference, and has a very good development prospect. In China, it can provide a scientific basis for the state and governments at all levels to make macro-decisions on grain production and planning, and to formulate correct policies on grain distribution, supply, storage and transportation and trade at home and abroad.