Cherry blossom and ginkgo leaf coloration phenology timing of China from 2009 to 2019 extracted from big data

This repository accompanies the manuscript <b>"Cherry blossom and ginkgo leaf coloration phenology dataset of China from 2009 to 2019 extracted from big data"</b> by Shenghong Wang, Haolong Liu, Xinyue Qin, Junhu Dai and <b>Jun Liu</b> *.Ground-based phenological observation data are the most accurate phenological monitoring data currently available. Making effective use of available information on social media to retrieve phenological data is of considerable value in alleviating the lack of phenological data in regions with missing observation sites. In this study, a logistic curve fitting method was developed to extract phenological data on specific species from social media data. After verifying the relationship between the site observation data and the temperature, timing data for two typical phenological phenomena in China, namely cherry blossom flowering in spring and ginkgo leaf coloration in autumn were reconstructed and published. The data availability is from 2010 to 2019 in 176 cities and 2009 to 2018 in 155 cities. This dataset is an effective supplement for existing phenological data, and this method also provides a reference for obtaining phenological data for specific species.The code for extract timing and data validation is available for download from the repository on GitHub at https://github.com/rebootcat/cherryGinkgoPhenology.

本仓库配套于王盛宏、刘浩龙、秦心玥、戴俊虎与**刘军*** 所撰写的论文《基于大数据提取的2009-2019年中国樱花与银杏叶色物候数据集》。地面物候观测数据是当前精度最高的物候监测数据。有效利用社交媒体公开信息获取物候数据，对于弥补观测站点缺失区域的物候数据缺口具有重要应用价值。本研究提出了一种逻辑斯蒂曲线拟合方法，可从社交媒体数据中提取特定物种的物候数据。在验证站点观测数据与气温的相关性后，本研究重构并发布了中国两种典型物候现象的物候期数据：春季樱花盛花期与秋季银杏叶变色期。本数据集的覆盖范围为：176座城市的2010-2019年数据，以及155座城市的2009-2018年数据。该数据集是现有物候数据的有效补充，本研究提出的方法也为特定物种的物候数据获取提供了参考。用于物候期提取与数据验证的代码可从本仓库的GitHub页面下载，链接为https://github.com/rebootcat/cherryGinkgoPhenology。