Scholarlink

<p>The human–nature relationship (HNR) has evolved profoundly over the past two centuries, shaped by socio-economic, cultural, and environmental factors. Despite cross-sectional studies identifying key influences, their longitudinal effects remain underexplored. This study models HNR trends from 1810 to 2020 using word frequency data as a proxy for the HNR and seven macro-level factors: natural disasters, scientific advancement, religiosity, industrialisation, humanism, urbanisation, and economic development. These data were analysed using a machine learning approach combining grid search and gradient descent-optimised weighted linear combination model, achieving excellent precision (RMSE = 0.0184). Urbanisation and economics emerged as dominant drivers of a HNR decline of 51.3% since 1850. While industrialisation’s importance diminished over time. The rise of humanism was identified as an increasingly significant influence, whereas science, natural disasters, and religiosity had relatively modest impacts. External validation using a modern dataset of nature connectedness across 52 nations demonstrated robust generalisability and utility across diverse contexts. The cross-validation of temporal consistency suggested utility for exploratory modelling of future scenarios of the HNR to 2050. The on-trend and recovery scenarios and factor weights highlight the importance of nature rich urbanisation and moderating increasingly human-centred values to ensure a closer HNR, thereby informing strategies that target the underlying causes of the environmental crises.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The human–nature relationship across two centuries: macro factor insights from a machine learning model

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月23日

查看全文
  • Miles Richardson

摘要

The human–nature relationship (HNR) has evolved profoundly over the past two centuries, shaped by socio-economic, cultural, and environmental factors. Despite cross-sectional studies identifying key influences, their longitudinal effects remain underexplored. This study models HNR trends from 1810 to 2020 using word frequency data as a proxy for the HNR and seven macro-level factors: natural disasters, scientific advancement, religiosity, industrialisation, humanism, urbanisation, and economic development. These data were analysed using a machine learning approach combining grid search and gradient descent-optimised weighted linear combination model, achieving excellent precision (RMSE = 0.0184). Urbanisation and economics emerged as dominant drivers of a HNR decline of 51.3% since 1850. While industrialisation’s importance diminished over time. The rise of humanism was identified as an increasingly significant influence, whereas science, natural disasters, and religiosity had relatively modest impacts. External validation using a modern dataset of nature connectedness across 52 nations demonstrated robust generalisability and utility across diverse contexts. The cross-validation of temporal consistency suggested utility for exploratory modelling of future scenarios of the HNR to 2050. The on-trend and recovery scenarios and factor weights highlight the importance of nature rich urbanisation and moderating increasingly human-centred values to ensure a closer HNR, thereby informing strategies that target the underlying causes of the environmental crises.