Map vector tile construction for arable land spatial connectivity analysis based on the Hadoop cloud platform

Abstract

The concentrated connection of arable land is one of the important indicators reflecting the quality of cultivated land, and large-scale arable land blocks are more conducive to agricultural mechanization operation, thereby improving the land use efficiency. However, the calculation of farmland connectivity is essentially a large-scale calculation of spatial vector data, especially for the national or global farmland patch data. This article proposes a framework for calculating farmland connectivity based on spatial vector map tiles and parallelizes the algorithm based on the Hadoop cloud platform. The framework is based on the tile pyramid model and uses the Douglas–Peucker algorithm to simplify the data to meet the needs of rapid display of large-scale data under multi-scale. The consistency and integrity of the front display of vector tiles are ensured using the setting tile buffer. Meanwhile, the parallelization of the vector tile construction algorithm is realized based on the MapReduce programming mode. Finally, the effectiveness and usability of this framework were verified through the calculation of patch connectivity on the tillage map. Experiments show that the algorithm can not only meet the rapid construction requirements of large-scale vector tile data but also support the cultivated land spatial connectivity analysis and greatly improve the efficiency of supporting data calculation.