干旱区研究

Arid Zone Research ›› 2019, Vol. 36 ›› Issue (3): 781-789.doi: 10.13866/j.azr.2019.03.30

• Others • Previous Articles    

Extraction of Crops in Oasis Based on GF-1/WFV Time Series

LIU Ya-qing1,2, WANG Lei1,2,3, ZHAO Xi-ni1,2, QU Xiang-ning1,2, XU Xing1,2, WANG Rui1,2   

  1. 1. Breeding Base for State Key Laboratory of Land Degradation and Ecosystem Restoration in Northwest China, Ningxia University, Yinchuan 750021,Ningxia,China;
    2. Key Laboratory for Restoration and Reconstruction of Degenerated Ecosystem in Northwest China under Ministry of Education, Ningxia University,Yinchuan 750021,Ningxia,China;
    3. International Institute for Earth System Science,Nanjing University,Nanjing 210093,Jiangsu,China
  • Received:2018-08-17 Revised:2018-09-20 Published:2025-10-18

PDF (PC)

0

Like

Abstract: The rapid extraction of regional crops is of great significance for agricultural production management,planting structure adjustment and optimization,and food security.The use of time series data of remote sensing image for extracting crops is an important means,and the time resolution and spatial resolution are constraints.The time resolution of medium spatial resolution of remote sensing data is significantly improved with the successful launch of the first satellite GF-1 of China High-resolution Earth Observation System.However,the seasonal rhythm features of crops differ greatly under different climate backgrounds.The Hetao Oasis is irrigated by the Yellow River.It has an ecological landscape with obvious microclimate effect.The extraction of remote sensing data of crops is typical and representative.The Hetao Oasis in Ningxia is selected as the study area to analyze and evaluate the applicability of GF-1 satellite data in extracting crops.According to the characteristics of crop planting in the study area,the four main crops including rice,corn,wheat and alfalfa were extracted.Firstly,the GF-1/WFV image was calculated by band,and the NDVI and NDWI time series of remote sensing data sets were obtained.The values of NDVI and NDWI of the main crops were extracted,and the time series curves of NDVI and NDWI were constructed.The time series data of remote sensing indexes of the main crops were analyze,the farming information was determined to extract the phases and thresholds of the crops,construct the decision tree and extract the crops.The time series data of natural grasslands,reed and grape and of other vegetation types were combined to establish the CART decision tree so as to verify the classification effect of the correlation time series and the decision tree constructed by the farming information.The results are as follows:① The morphological characteristics of time series curves were different from different farming ways,and the NDVI time series data could be used to describe the growth status of crops in different periods; the NDWI time series data obtained from GF-1/WFV could be used to describe the soil moisture status of different crops and reflect the different irrigation systems of different crops in the study area.The NDVI time series and the NDWI time series could be used to accurately express the surface dynamic change information of the surrounding areas of the study area,correlate the crop time series with farming information,and effectively recognize different crops;② Compared with the CART decision tree,the decision tree of classification phase,classification threshold and classification accuracy constructed by combining different crop farming information and main crop curve features were similar.These indicated that the farming ways were different from different crops in the study area,and the time series of corresponding NDVI and NDWI values were significantly different.

Key words: crops, cropping pattern, GF-1/WFV time series data, decision tree, remote sensing mapping, oasis, Ningxia