Search

Shelterbelt width not only influences protection effectiveness, but also represents a variable in a porosity model. Following extensive planting of shelterbelts globally, it is important to accurately identify shelterbelt width to evaluate its effectiveness on the regional scale. However, obtaining such an inventory is challenging given the narrow linear characteristic and extensive distribution. The development of high-resolution remote sensing monitoring offers the possibility of accurate identification of shelterbelt width. This study developed a method for identification of shelterbelt width using 3-m spatial resolution JL1–02A remote sensing imagery. The primary principles of the method are as follows: (1) generate two-dimensional waveforms of shelterbelts using remote sensing images; (2) recognize intact parts of shelterbelts from wave features in the direction of belt orientation; (3) determine the belt edge and pixels related to belt width through neighboring pixel analysis in the direction of belt width; and (4) establish an algorithm for shelterbelt width identification through analysis of its relationship with edge and inner pixels. Using the proposed method, shelterbelt width was identified in the study area in Jilin Province (Northeast China), and the accuracy of the measurement was analyzed using a 0.75-m resolution fusion image (sample number: 140). Results showed that the coefficient of determination (R2) was 0.93, root mean square error was 2.18 m, mean absolute error was 1.76 m, maximum error was 5.11 m, and minimum error was 0.01 m, which verified the accuracy and reliability of the method for shelterbelt width identification. The result could be used to determine shelterbelt structure (e.g., number of rows), thereby providing the possibility for remote simulation of shelterbelt porosity, which was previously only measured from the ground-level approach. The proposed method could help promote accurate monitoring and management of shelterbelt structure and function. [ABSTRACT FROM AUTHOR]