Credible historical land use/cover data are very important for past global change research. This study generates a set of integrated reconstruction methods based on multisource data and produces a new set of improved historical cropland data sets in Europe over the past 200 years. For AD 2000, FAO data, existing research results and statistical data are integrated. For AD 1900, a method of integrating two sets of independent historical agricultural land data by correction and validation and supplemented by other historical cropland data are developed. For AD 1850 and 1800, a methodological scheme of diversified proxy integrative technology and methods based on multisource data is constructed. In this new data set, quantitative reconstructions for AD 1900, 1850 and 1800 are improved to account for 100, 78 and 57% of all European countries, respectively. The reconstruction results show that each region in Europe has been in different stage of historical agricultural development. More than 86% of the countries’ cropland area and its proportion peaked in AD 1900 or did not exceed the data for AD 2000. Specifically, a high reclamation zone gradually formed from France to Ukraine, in which every country’s cropland fraction was ≥40% during AD 1800–1900. From AD 1900–2000, the highly cultivated region contracted, and the centre of higher cropland proportions shifted to eastern Europe and Poland, Czechia and Hungary. The cropland area was systematically underestimated by HYDE3.2, with a relative difference ratio of −20 to −30% between HYDE3.2 and this study. Historical empirical data were used for only 32% of countries in HYDE3.2. This method of multiproxy integrated reconstruction is applicable to other regions of the world and it would be worth attempting to apply it to earlier historical European cropland data sets in the future.

中文翻译：

---

使用综合多源数据重建自公元 1800 年以来欧洲国家的耕地变化

可靠的历史土地利用/覆盖数据对于过去的全球变化研究非常重要。本研究生成了一套基于多源数据的综合重建方法，并生成了一套新的改进的欧洲过去 200 年历史农田数据集。对于 AD 2000，整合了 FAO 数据、现有研究结果和统计数据。针对公元1900年，开发了一种通过校正和验证整合两组独立历史农地数据并辅以其他历史农田数据的方法。针对公元1850年和1800年，构建了基于多源数据的多样化代理综合技术和方法的方法论方案。在这个新数据集中，AD 1900、1850 和 1800 的定量重建得到改进，分别占所有欧洲国家的 100%、78% 和 57%，分别。重建结果表明，欧洲各个地区都处于不同的历史农业发展阶段。86%以上的国家耕地面积及其比例在公元1900年达到顶峰或未超过公元2000年的数据。具体而言，从法国到乌克兰逐渐形成高垦区，其中各国耕地比例≥40%公元 1800-1900 年间。从公元 1900 年到 2000 年，高度耕种的区域缩小，较高耕地比例的中心转移到东欧和波兰、捷克和匈牙利。HYDE3.2系统地低估了耕地面积，HYDE3.2与本研究之间的相对差异率为-20%至-30%。HYDE3.2 中仅 32% 的国家使用了历史经验数据。