Lightweight Wrappers for Adapting Time Series Foundation Models to Regional Drought Forecasting

📄 arXiv: 2607.17511v1 📥 PDF

作者: Wentao Gao, Jiuyong Li, Lin Liu, Thuc Duy Le, Jixue Liu, Yanchang Zhao, Yun Chen

分类: cs.LG

发布日期: 2026-07-20

备注: 12 Pages, KDD2026 Accpeted Paper(Poster)


💡 一句话要点

提出轻量级适配框架以解决区域干旱预测问题

🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 时间序列预测 干旱预测 模型适配 轻量级框架 多分辨率视图 块重采样 机器学习 气候建模

📋 核心要点

  1. 现有时间序列预测方法在区域气候预测中面临模型权重专有、本地数据不足和计算资源限制等挑战。
  2. 本文提出了一种轻量级黑箱适配框架,通过SMR²和MBB两个包装器在推理时增强冻结的时间序列基础模型。
  3. 在南澳大利亚的实验中,该框架在多个基础模型上均实现了显著的预测性能提升,MSE减少了26%。

📝 摘要(中文)

大型时间序列基础模型(TSFMs)在多个领域展现出强大的零-shot 预测能力。然而,它们在区域气候预测中的应用面临实际挑战:模型权重通常是专有的,本地训练记录有限,计算预算受限,使得传统微调方法不可行。为了解决这些限制,本文提出了一种轻量级的黑箱适配框架,通过两个即插即用的包装器(SMR²和MBB)在推理时增强冻结的TSFMs。SMR²通过多分辨率时间视图分解输入,学习特定步幅的残差修正,并自适应集成成单一预测;MBB则通过块重采样保持时间依赖性,并对时间一致的残差扰动进行集成以稳定点预测。该框架在南澳大利亚多个站点的标准化降水蒸散指数预测中,显著提高了预测性能,平均均方误差(MSE)减少了26%。

🔬 方法详解

问题定义:本文旨在解决大型时间序列基础模型在区域干旱预测中的应用问题,现有方法由于模型权重专有和本地数据不足,导致传统微调不可行。

核心思路:提出一种轻量级的黑箱适配框架,利用推理时的集成方法增强冻结的TSFMs,避免对模型权重的直接修改。

技术框架:框架包含两个主要模块:SMR²和MBB。SMR²通过多分辨率视图分解输入,学习特定步幅的残差修正;MBB则通过块重采样保持时间依赖性,进行残差扰动的集成。

关键创新:最重要的创新在于通过推理时的集成方法实现适配,而非传统的模型微调,确保了适配过程的轻量级和高效性。

关键设计:SMR²模块设计了多分辨率视图和自适应集成机制,MBB模块则采用块重采样技术,确保时间依赖性得到保留,所有适配均在推理阶段完成。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果表明,提出的框架在南澳大利亚多个站点的标准化降水蒸散指数预测中,较对应的冻结基础模型实现了最高26%的均方误差(MSE)降低,展示了其在实际应用中的有效性和优势。

🎯 应用场景

该研究的潜在应用领域包括区域气候预测、农业管理和水资源管理等。通过提供一种高效的预测工具,能够在资源受限的环境中实现更准确的干旱预测,进而为决策提供支持,具有重要的实际价值和未来影响。

📄 摘要(原文)

Large \emph{Time Series Foundation Models} (TSFMs) demonstrate strong zero-shot forecasting capabilities across diverse domains. However, their application to regional climate forecasting faces practical challenges: model weights are often proprietary, local training records are limited, and computational budgets are constrained, making traditional fine-tuning approaches infeasible. To address these constraints, we introduce a lightweight, black-box adaptation framework (requiring no access to backbone parameters and no backbone fine-tuning) that enhances frozen TSFMs at inference time through two plug-and-play wrappers: \textbf{SMR\textsuperscript{2}} (Stationarity aware multi-resolution Residual), which decomposes the input into multi-resolution temporal views, learns stride specific residual corrections that capture regional dynamics, then adaptively ensembles them into a single forecast, and \textbf{MBB} (Moving Block Bootstrap), which preserves temporal dependencies through block resampling and ensembles over temporally coherent residual perturbations to stabilize the point forecast. Both wrappers instantiate the same bagging style principle: they build diverse views of the input or its residuals, forecast each with the same frozen backbone, and aggregate, so all adaptation comes from inference time ensembling rather than any weight update. Evaluated on one month ahead Standardized Precipitation Evapotranspiration Index (SPEI) prediction across multiple sites in South Australia, our framework consistently improves forecasting performance across several backbone models, demonstrating up to 26\% mean squared error (MSE) reduction over the corresponding frozen backbone while enabling practical deployment in resource constrained regional forecasting systems.