Study Of Universal Eco-Meteorological And Agro-Meteorological Automatic Observation Method

Eco-meteorological and agro-metrological observation focuses on the climate condition, environment, vegetation status, human activity and their inter-relationships. Therefore, it is vastly important not only in order to optimize fanning, but also in order to research ecological variation. The objects of eco-meteorological and agro-meteorological observation usually include meteorology factors such as sunshine, radiation, air temperature, precipitation, soil temperature, and soil moisture, the crop growth status such as height, density, leaf area index, growth cycle, and the phenological indicators such as germination or the flowering of specific plants, migratory birds movement, and thunder. Among these indexes, only some of them can be measured by instrument, while most information depends on manual sampling and estimation. These traditional observation methods are not only labor-intensive but also inaccurate, as the results are often subjective. In order to have a thoroughly diverse data set for modern agriculture and agro-ecological environment monitoring, a new technology and method of observation is required.Based on an examination of today's automatic observation systems, a universal automatic system for eco-meteorology and agro-meteorology observation was designed, testing stations with said systems were built, and methods in data analysis and use were begun. Our purpose was to find a new technology which not only met the requirements for unmanned continuous observation, but also reflected the characteristics of eco-meteorology and agro-meteorology.The main contents and results are presented as follows:1. A universal eco-meteorological and agro-meteorological automatic observation system was designed and developed. Considering the fact that most observation needs to be taken in scattered locations with diverse observational factors, this system was designed with distributed system architecture, functional modularity, and Virtual Instrument technology, which make it not only stable, flexible and scalable, but also provides various information. The system was designed with a lower computer and a upper computer. At each station, the main controller of the lower computer is a data collector which is composed of a cFP and digital I/O models. Combined with different types of sensors and wireless communication equipment, it can collect, process, save, and transport different types of data in different frequencies which are uploaded to a service center automatically. The upper computer in the service center is responsible for data analysis, output generation and saving. For the first time, the system integrates a digital image/audio collector, which can record the change in the surroundings either over time or in real time. Furthermore, a solar and wind electrical generator, tamper resistant equipment, and real time status indicator enable continual power and security for the unmanned automatic stations, making them especially suitable for long term observation in uninhabited areas. Therefore, this makes the remote observation of crops and phenology possible.2. In order to test the practicality and reliability of this system, testing stations were built, adjusted and run in Guangxi, Shandong, Qinghai, Henan and Jiangxi. The testing was divided into two periods. The first period was for four seasons in Guangxi to examine its ability to handle the stress from continual operation in different weather conditions. In the second period, two-month tests were done in each of the other four stations, focusing on the stability and accuracy of the observation data. The results showed that the automatic observation system is well designed for the agro-ecosystem and was rarely affected by either geography circumstance or weather conditions. In fact, the average data loss rate was less than 0.93%. Moreover, the cost in construction and operation of the system is much less than a normal observation station.3. An automatic observation method and an identification method of crop growing season was designed. In this research, parallel observations were taken through the whole growth cycle of both rice and corn to compare the automatic and manual methods. First, the automated observation system obtained digital images of canopy from different angles. The growing period characteristic index to be used by automatic observation were proposed. Those indexes, such as specific color, change of shape, and organ appearance can be identified by a color-based image process and morphology recognition technology. This is not only an effective way for crop growth monitoring, making the observation much more convenient and objective, but also offers suitable indexes for crop growth status determination.4. By using object-based image analyses based on the spectral characteristics of objects in different periods, a means to automatically classify objects and an algorithm to quantify percent ground cover were studied. The analysis showed a high accuracy compared to the estimated coverage conducted by the point-by-point calculation method, approximately 98%, with a correlation coefficient between automatic estimation and manual calculation of 0.992 (p<0.0001). Moreover, the resulting slopes reflected the growth curve variation of the green leaves during the growth cycle. The percent ground cover calculated by this method had a significant exponential correlation with the measured leaf area index. Furthermore, by combining the automatic system images with aerial remote sensing, the mixed vegetation coverage of the fields around the Guangxi station were calculated. This method offers a process of mixed pixel interpretation in remote sensing.5. To apply the system and the methods, meteorological factors, soil data, and two types of grassland coverage were collected from the Qinghai automatic station through July to October 2010. An analysis was then taken on the effect of environmental factors to grass growth based on these data. The vegetation index was also calculated from the coverage rate, and was compared with satellite NDVI. Results showed that:(1) The difference on the growing period and the response to environmental change between different types of grass can be reflected through their coverage curves. The application of the automatic observation process in ecological monitoring and research can sensitively reflect the relation between vegetation and environmental factors. It provides detailed information to study area partition according to vegetation type or circumstance condition, and therefore makes it possible to fully consider the driving forces of vegetation change. (2) Heat condition and water status are the main factors for grass growth in that area. The average air temperature, minimum air temperature and topsoil temperature are the key factors which control the length of growing season. While the average air temperature was above 10℃and topsoil temperature was above 15℃, the grassland coverage had a closer correlation with water pressure and soil moisture. (3) Due to satellite sensor distortion, cloud cover, and data processing method limitations, satellite NDVI might be biased. Therefore, the land coverage data obtained from the automatic observation systems could be used in satellite data correction.