Climate variability, as an element of uncertainty in water management, affects community, sectoral, and individual decision-making. Long-range prediction models are tools that offer the potential for integration and joint analysis with the hydrological, hydrodynamic, and management response of the socio-ecological systems to which they are linked. The main objective of this article is to present a seasonal climate prediction model, the open-source algorithm SIE-Climate, whose application consists of three phases (exploration, development, and evaluation), and to describe its application to the Lake Sochagota socio-ecological system (Paipa, Boyacá, Colombia). The K-nearest neighbours method is used when defining a target matrix that represents and integrates macro- and micro-climatic phenomena (Oceanic Niño Index, local temperature, and local rainfall) to identify periods of similar climatic behaviour. Considering a 1-year horizon and management purposes the tool is calibrated and validated in periods with and without climatic anomalies (2000–2018), giving reliable adjustment results (RSME:4.86; R²: 0.95; PBIAS: −8.89%; EFF: 0.85). SIE-Climate can be adapted to various contexts, variables of interest, and temporal and spatial scales, with an appropriate technological and computational cost for regional water management.

中文翻译：

---

SIE-Climate：在管理社会生态系统中预测当地气候变化的方法和技术工具

气候变率是水资源管理中不确定性的一个因素，影响社区、部门和个人的决策。长期预测模型是一种工具，可提供与与其相关的社会生态系统的水文、水动力和管理响应进行整合和联合分析的潜力。本文的主要目的是提出一个季节性气候预测模型，即开源算法 SIE-Climate，其应用包括三个阶段（探索、开发和评估），并描述其在索查戈塔湖社会-生态系统（Paipa，Boyacá，哥伦比亚）。在定义表示和整合宏观和微观气候现象（海洋 Niño 指数、当地温度、和当地降雨量）来确定类似气候行为的时期。考虑到 1 年的范围和管理目的，该工具在有和没有气候异常的时期（2000-2018 年）进行校准和验证，给出可靠的调整结果（RSME：4.86；R ²：0.95；PBIAS：-8.89%；EFF：0.85）。SIE-Climate 可以适应各种环境、感兴趣的变量以及时间和空间尺度，并为区域水资源管理提供适当的技术和计算成本。