Geoskill: 撂荒地检测
Detecting Long-term Abandoned Cropland Using NDVI Time Series Data
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: Geoskill: 撂荒地检测. It is used for: Detecting Long-term Abandoned Cropland Using NDVI Time Series Data Full Skill content: https://321skill.com/skills/abandoned-farmland-detector/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
This skill focuses on detecting cropland suspected of being abandoned for multiple consecutive years by analyzing Normalized Difference Vegetation Index (NDVI) time series data. It addresses the challenge in agricultural monitoring and land resource management where manual tracking of cropland use status over large areas and extended periods is difficult.
To use it, users need to provide remote sensing imagery or NDVI time series data for the target area. The skill analyzes multi-year data to identify cropland patches where the NDVI variation pattern over several consecutive years does not align with normal cultivation cycles, and outputs the location and extent of suspected abandoned plots.
This skill is particularly suitable for agricultural management departments, land resource survey units, agricultural researchers, and analysts involved in agricultural insurance or investment assessment. It helps them efficiently screen for abandoned cropland, generate field verification checklists, or assess regional cultivation continuity.
Before use, it is recommended to ensure the temporal coverage and quality of the NDVI data; typically, at least 3–5 years of continuous data are required to improve detection accuracy. Additionally, as the detection results serve as clues for suspected abandonment, they should be verified through field checks or higher-resolution imagery to rule out false positives caused by crop rotation, fallowing, or data noise.
Key Features
Unlike general vegetation change monitoring, this skill is specifically designed for cropland abandonment scenarios. It identifies the persistence of "no cultivation signals" by analyzing NDVI time series characteristics over multiple consecutive years, rather than focusing on single or short-term vegetation cover changes. It targets the continuous interruption of agricultural activity, and its output directly provides a list of suspected abandoned cropland, aligning more closely with the practical needs of agricultural management and land supervision.
Limitations
It is not suitable for monitoring vegetation changes in non-cropland areas (e.g., forests, grasslands), and it requires high-quality input NDVI data with complete time series and effective cloud contamination handling.
FAQ
How many years of data are needed to effectively detect abandoned cropland?
It is recommended to provide at least 3–5 years of continuous NDVI time series data. The longer the time span, the more reliable the detection results.
How accurate are the detected abandoned cropland results?
The results indicate "suspected" abandoned cropland based on remote sensing signal analysis. Final confirmation is recommended through field verification or high-resolution imagery.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/abandoned-farmland-detector/raw/index.md to read the original Skill definition (Markdown format) for Geoskill: 撂荒地检测, and install it according to the instructions.
Raw Markdown URL for AI: /skills/abandoned-farmland-detector/raw/index.md