数据与计算发展前沿 ›› 2026, Vol. 8 ›› Issue (5): 48-60.

CSTR: 32002.14.jfdc.CN10-1649/TP.2026.05.004

doi: 10.11871/jfdc.issn.2096-742X.2026.05.004

• 专刊：“东方”超级计算系统，助力青年科学家科技创新 • 上一篇    下一篇

基于多源数据的耕地生态要素评价与聚类分区保护研究——以海伦农场为例

康美琪1(),王崇倡1,于灵雪2,3,杨月稳4,闫卓冉2,3,*()   

  1. 1 辽宁工程技术大学， 测绘与地理科学学院， 辽宁 阜新 123000
    2 中国科学院东北地理与农业生态研究所， 黑土地保护与利用重点实验室， 吉林 长春 130102
    3 吉林省生态环境遥感大数据重点实验室， 吉林 长春 130102
    4 西南林业大学， 经济管理学院， 云南 昆明 650224
  • 收稿日期:2026-04-23 出版日期:2026-10-20 发布日期:2026-10-20
  • 通讯作者: *闫卓冉（E-mail：yanzhuoran@iga.ac.cn）
  • 作者简介:康美琪，辽宁工程技术大学，硕士研究生，主要研究方向为基于遥感与空间分析的耕地多维生态要素监测。
    本文主要承担工作为数据处理、实验设计与论文的撰写与修改。
    KANG Meiqi is a master’s student at Liaoning Technical University. Her research interests include monitoring of multi-dimensional ecological elements of cultivated land based on remote sensing and spatial analysis.
    In this paper, she is mainly responsible for data processing, experimental design, as well as manuscript writing and revision. E-mail:17741226026@163.com|闫卓冉，中国科学院东北地理与农业生态研究所，博士，助理研究员，主要研究方向为生态系统监测与模拟。
    本文主要承担工作为论文框架构建与内容修改。
    YAN Zhuoran, Ph.D., is an assistant researcher at the Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences. His research interests include ecosystem monitoring and simulation.
    In this paper, he is mainly responsible for manuscript framework construction and content revision. E-mail:yanzhuoran@iga.ac.cn
  • 基金资助:
    国家重点研发计划(2024YFD1501600);“东方之星”青年科学家创新基金(DFZX202517)

Evaluation of Cultivated Land Ecological Elements and Clustering Zoning Protection Based on Multi-Source Data: A Case Study of Hailun Farm

KANG Meiqi1(),WANG Chongchang1,YU Lingxue2,3,YANG Yuewen4,YAN Zhuoran2,3,*()   

  1. 1 College of Geomatics, Liaoning Technical University, Fuxin, Liaoning 123000, China
    2 Key Laboratory of Mollisols Agroecology, Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun, Jilin 130102, China
    3 Jilin Provincial Key Laboratory of Ecological Environment Remote Sensing Big Data, Changchun, Jilin 130102, China
    4 College of Economics and Management, Southwest Forestry University, Kunming, Yunnan 650224, China
  • Received:2026-04-23 Online:2026-10-20 Published:2026-10-20

RichHTML

1

PDF

6

摘要：

【目的】科学认知耕地多维生态要素的空间异质性特征，对破解农业粮食生产与其他生态服务间的权衡关系具有重要意义。【方法】本研究以位于黑龙江省的海伦农场为研究区域，基于多源数据构建土壤-作物-景观多维生态评价体系，应用冷热点分析与局部双变量空间自相关方法揭示多维生态要素的空间分异格局与相互作用关系，并采用自组织映射神经网络划定耕地生态治理综合分区。【结果】结果表明：（1）研究区耕地多维生态要素等级分化明显，其中土壤环境以中等级为主并多局部集聚于研究区西部地区，作物长势高等级地块数量最多且呈现空间均质化特征，景观生态低等级地块分布最广且空间破碎化风险较强。（2）研究区内景观与土壤维度的权衡与协同地块数量相对一致，而作物长势与其他生态要素的空间权衡关系更为突出。当前集约化农业模式导致作物优良长势的维持过度依赖人为投入，割裂的农业生态系统内部联系对海伦农场地力的永续利用构成潜在威胁。（3）遵循聚类分区组内差异最小、组间差异最大原则，将研究区划分为综合整治区、生态优化区、提质增效区及生态保育区四类分区，不同分区应结合要素短板进行针对性环境治理与生态养护。【结论】研究可为海伦农场及其他农业集约化地区的耕地生态多维评估与差异化修复策略提供科学参考。

关键词: 耕地保护, 生态要素, 空间聚类, 人工神经网络, 遥感评价

Abstract:

[Objective] A scientific understanding of the spatial heterogeneity of multidimensional ecological elements in cultivated land is of great significance for disentangling the trade-offs between agricultural grain production and other ecological services. [Methods] Taking Hailun Farm in Heilongjiang Province as the study area, this research constructed a soil-crop-landscape multidimensional ecological evaluation system based on multi-source data. Hot and cold spot analysis and local bivariate spatial autocorrelation methods were applied to reveal the spatial differentiation patterns and interactions of multidimensional ecological elements, and a self-organizing map (SOM) neural network was employed to delineate comprehensive zones for cultivated land ecological management. [Results] The results indicate that: (1) The grades of multidimensional ecological elements of cultivated land in the study area exhibit distinct differentiation. Specifically, the soil environment is predominantly of medium grade and mostly locally clustered in the western part of the study area; plots with high-grade crop growth are the most numerous and show spatial homogeneity; and plots with low-grade landscape ecology are the most widely distributed, presenting a strong risk of spatial fragmentation. (2) The numbers of plots showing trade-offs and synergies between the landscape and soil dimensions are relatively balanced, whereas the spatial trade-offs between crop growth and other ecological elements are more prominent. The current intensive agricultural model leads to an over-reliance on artificial inputs to maintain excellent crop growth, and the presently fragmented internal connections within the agroecosystem pose a potential threat to the sustainable utilization of soil fertility in Hailun Farm. (3) Following the clustering principle of minimizing intra-group variance and maximizing inter-group variance, the study area is divided into four types of zones: comprehensive remediation zone, ecological optimization zone, quality and efficiency enhancement zone, and ecological conservation zone. Targeted environmental governance and ecological conservation should be implemented in different zones based on their respective limiting factors. [Conclusions] This study provides a scientific reference for the multidimensional ecological assessment and differentiated restoration strategies of cultivated land in Hailun Farm and other intensive agricultural areas.

Key words: cultivated land protection, ecological factors, spatial clustering, artificial neural networks, remote sensing assessment