A seamless economical feature extraction method using Landsat time series data

Abstract

Listen

Translate article

Regional economic development describes the total social and economic activities in a given time and space. An objective understanding of the real regional economy is beneficial for healthy, sustainable societal development. Generally speaking, the understanding of the regional economy is mainly based on social surveying, which incurs time and energy costs and lacks objectivity. Therefore, this study proposes a seamless economical feature extraction method using the advantages of Landsat time series based on the morphologic changes of the earth’s surface caused by regional economic development. First, the land-use/cover changes of the earth’s surface were collected using Landsat time series; second, the correlations between land-use types and regional economic indices were analyzed and the optimal sensitive factors were selected. Third, a regional economic development model was constructed from the perspective of the land-use/cover change observed by remote sensing technology. Finally, the accuracy was evaluated in order to assess the validity and applicability of the model. The Zhoushan Islands of China were chosen as the research area for the verification experiment. From the results, the construction land is the most significant sensitive factor that correlates closely with various economic indices, and its correlation coefficients R with gross domestic product (GDP), value-added of primary industry (VPI), value-added of secondary industry (VSI), and value-added of tertiary industry (VTI) were 0.9591, 0.9390, 0.9546, and 0.9573, respectively. The regional economic development model constructed is simple, clear, and highly accurate; the determination coefficient R2 was 0.9884. This study opens up unique opportunities for the objective, seamless understanding of regional economic development from the perspective of land-use/cover change using Landsat time series, as well as the correction of economic survey data, both with a high degree of accuracy.