Mapping Climate Connectivity and Hot Spots over Türkiye Using a Station-Based Network Framework

Assessing climate-change impacts solely from local trends may be misleading, as large-scale teleconnections transmit anomalies across regions and may reorganize spatial risk, potentially creating dynamically influential “hot spots.” This study presents a station-based climate-network (“climate grid”) framework to quantify spatiotemporal connectivity and its long term evolution across Türkiye. The network is built from 221 in situ stations operated by the Turkish General Directorate of Meteorology and is based on 1960–2024 average daily temperature and total daily precipitation observations.

All computations are performed in Python environment using equations formulated for climate-grid construction and are parallelized to handle to the full set of pairwise station calculations. Prior to network construction, station time series are deseasonalized and standardized to anomalies to ensure cross-regime comparability across Türkiye’s diverse climatic regimes and topography, in line with standard climate-network methodologies. A Monte Carlo permutation procedure is applied to test link significance and to define objective sparsification thresholds, minimizing spurious connections.

Hot spots are identifies through network centrality analysis, emphasizing degree and its directional components (in-degree/out-degree) to classify stations that mainly drive regional variability (“sources”) from those that mainly respond to it (“sinks”). This is consistent with prior degree-based hot-spot detection that combine degree-based hot-spot mapping with nonparametric trend testing (e.g., Mann–Kendall) to evaluate changes under anthropogenic forcing. Guided by recent network studies on how extremes propagate through networks and how drought conditions synchronize directionally, the framework supports to track spatiotemporal connectivity through time and to identify regions with potential cascading behavior.

The resulting climate grid is separated into two complementary components: (i) a focal-station network that summarizes the links from a station to its surrounding stations, and (ii) a reciprocal network describing the surroundings’ connections back to the focal station. Explicitly representing both outward and inward connectivity provides a directional interpretation of climate coupling and allows stations to be characterized as potential “sources” versus “sinks” of climate influence. Hot spots are then identified using network centrality measures. This allows us to map influential and sensitive locations across Türkiye, assess how connectivity patterns shift over time, and help prioritize monitoring and adaptation actions under increasing climate variability and extremes. The results are presented and discussed in terms of national-scale connectivity patterns, hot-spot persistence, and emerging shifts through time.

Keywords: climate networks; spatiotemporal connectivity; station-based grid; Türkiye; time-lagged dependence; hot spots.