Dataset for TAE-ConvLSTM-based Ecological Quality Forecasting on the Qinghai–Tibet Plateau

This repository contains the full dataset and model code supporting the paper "Temperature-driven Ecological Dynamics under Future Climate Change across the Qinghai–Tibet Plateau", submitted to *Environmental Modelling & Software*. The dataset includes: 1. **CMIP6 climate forcing data** (SSP1–2.6, SSP2–4.5, SSP3–7.0, SSP5–8.5) used as model inputs. 2. **Remote Sensing Ecological Index (RSEI) layers** derived from Landsat and MODIS imagery. 3. **Processed ecological quality indicators** and spatial grids used for model training. 4. **Trained TAE-ConvLSTM model weights and prediction results (2015–2100)**. The data and code are organized into two main directories: 1. origin_data/ Original input data used in the analysis. It includes: boundary_data/ — Geospatial boundary files used to define the study area. gcm_data/ — Climate projection data obtained from CMIP6 models. obs_data/ — Observational data used for analysis and validation. 2. processed_data/ Processed results and model outputs used in the study. It includes: step1、OBS/ — Processed observational datasets. step2、RSEI/ — Ecological index results generated from remote sensing data. step3、GCM/ — Processed or downscaled climate model data. step4、deeplearning/ — Deep learning–related files, including results. metric_result/ — Evaluation metric. All files are openly shared under the Creative Commons Attribution 4.0 International License (CC BY 4.0).本仓库包含了提交至《环境建模与软件》期刊的论文《青藏高原未来气候变化下温度驱动的生态动力学》所需的完整数据集与模型代码。 该数据集包含以下内容： 1. **第六次耦合模式比较计划（CMIP6）气候强迫数据**（涵盖SSP1–2.6、SSP2–4.5、SSP3–7.0、SSP5–8.5四种共享社会经济路径情景），作为模型输入使用。 2. **遥感生态指数（Remote Sensing Ecological Index, RSEI）图层**，基于Landsat与MODIS影像反演得到。 3. **经预处理的生态质量指标与空间格网数据**，用于模型训练。 4. **已训练完成的TAE-ConvLSTM模型权重及2015–2100年的预测结果**。 数据与代码分为两个主要目录： 1. `origin_data/`：分析所用的原始输入数据，其中包含： - `boundary_data/`：用于划定研究区域的地理空间边界文件。 - `gcm_data/`：从CMIP6模型获取的气候预估数据。 - `obs_data/`：用于分析与验证的观测数据。 2. `processed_data/`：本研究使用的预处理结果与模型输出文件，其中包含： - `step1/OBS/`：经预处理的观测数据集。 - `step2/RSEI/`：基于遥感数据生成的生态指数结果。 - `step3/GCM/`：经预处理或降尺度处理的气候模型数据。 - `step4/deeplearning/`：深度学习相关文件（含实验结果）。 - `metric_result/`：评估指标文件。 所有数据与代码均采用知识共享署名4.0国际许可协议（CC BY 4.0）公开共享。