Regional Applicability Analysis Of GF-1 Data In The Monitoring Of Crop Planting Area

Selected Hubei province as the study area,the regional applicability of Gaofen-1 satellite data in remote sensing monitoring of winter wheat,rape,early rice,and middle rice,cotton,corn and soybean staple crop planting area was analyzed from the aspects of topography,planting structure,the monitoring temporal,spectral deploying,monitoring accuracy.Through the research and analysis in this paper,the following conclusions are obtained:(1)Gaofen-1 satellite image data in Jianghan Plain of Hubei province,along the Yangtze River delta plain,mountainous areas of western Hubei province,in North hillock area in Hubei province and in the eastern hilly and area of winter wheat,rape and other remote sensing monitoring have better applicability.Among them,in North hillock area of Hubei Province and plain area have the better effect,and mountainous and hilly region stake second place.Remote sensing monitoring of winter wheat and rape,the image of the best monitoring time is selected,which is generally in late 3,and the image data of single time phase can be used to achieve a high level of accuracy.The monitoring accuracy can meet the precision requirements of large scale remote sensing monitoring,but it is incompetence for the requirements of small scale remote sensing monitoring precision.(2)Gaofen-1 satellite image data for remote sensing monitoring of early rice,different regions,the best monitoring temporal are also different,in Hubei plain area,single season rice general in June began planting,and rice planting intervals greater mutual interference effects,the best monitoring time in early May.For the earlier and the larger time span in southeast hilly area of Hubei,early rice and single season rice and other objects in different phase still exist the problems"different objects which have thesame spectrum and the same objects have different spectrum",which need continuous multi remote sensing data in early to late May or early June to extract the rice planting information.(3)The PMS sensor has higher image resolution than the WFV sensor,and the image can reflect the boundary of the object more clearly,and it is relatively accurate according to the target crop area.Regardless of is in plain area and hilly area,the unsupervised classification interpretation result of random point to verify the accuracy of solution are improved significantly,but ground sampling to verify the accuracy in the plain area did not raise too much than WFV sensor,and hilly area increase more.However,PMS image sensor width is small,the coverage is small,so it is difficult to meet the demand of large area crop remote sensing monitoring.(4)For the remote sensing monitoring of rice,cotton,corn and soybean crops in Hubei province,which the planting structure is relatively simple,and the areas with less variety of crops,the remote sensing monitoring effect is good by using GF-1 data,but these regions in Hubei province are rarely;For the complex planting structure and field(ground)block more broken,such as in the eastern hilly,and intercrop crops at the same time in Jianghan Plain,the remote sensing monitoring effect is poor based GF-1 data.The main reason is that the GF-1 image only have near infrared,red,green and blue four bands,and lack of SWIR band which is very sensitive to the physiological and ecological characteristics of crops,resulting in GF-1 image classification often appear in mixed phenomenon.The characteristic bands,resulting in Gaofen-1 image classification often appear in mixed phenomenon.In the process of agricultural remote sensing monitoring of these summer crops,only the score,image data is difficult to meet the demand for monitoring,need to adopt with the SWIR bands of other satellite image data,such as landsat8 oli data,can improve the accuracy of remote sensing monitoring and satisfy the monitoring requirements.