Seasonal climate forecasting has emerged as a critical tool for proactive decision-making in water resources management, agricultural planning, and climate risk reduction, particularly in regions characterized by high climatic variability and water scarcity. Among the operational seasonal prediction systems, the Canadian Seasonal to Interannual Prediction System (CanSIPS), developed by the Environment and Climate Change Canada (ECCC), has demonstrated notable skill in capturing large-scale climate variability, particularly in extratropical regions. However, despite its operational implementation in Canada and comprehensive evaluation across North America, the performance of this forecasting system has never been systematically and specifically assessed for Iran—a country with complex topography, diverse climate regimes, and significant vulnerability to hydroclimatic extremes.

Existing studies conducted in Iran have examined the ECCC/CanSIPS system only indirectly, as a member of the North American Multi-Model Ensemble (NMME), and have reported aggregated results across all models without isolating CanSIPS performance. Moreover, the validation approaches in these studies have predominantly relied on gridded reanalysis products (e.g., CRU, PERSIANN-CDR, GPCC) rather than ground-based synoptic observations. This is particularly problematic in arid and mountainous regions, where reanalysis datasets suffer from large uncertainties due to sparse observational networks and complex orographic effects. Consequently, evaluating model outputs against such products essentially represents a "comparison of two estimates" rather than a genuine validation against direct measurements.

A further limitation of previous research is the disproportionate focus on mean temperature (T2m), while daily extreme temperatures—specifically minimum (Tmin) and maximum (Tmax)—have received minimal attention. These variables are fundamental to assessing frost risk, heat stress, growing degree-days, evaporative demand, and crop water requirements, all of which are critical for agricultural planning and water resource management in Iran's semi-arid and arid environments. Despite Iran's pronounced climatic diversity, encompassing Caspian humid, mountainous cold, central arid, and southern coastal regimes, no systematic, region-specific analysis of forecast skill has been conducted, particularly in relation to the dominant teleconnection patterns (ENSO and NAO) that modulate seasonal predictability in these latitudes. Additionally, the operational value of seasonal forecasts is inherently tied to lead time and initialization month; however, most studies in Iran have reported aggregated skill metrics without analyzing skill decay across successive lead times or evaluating the sensitivity of performance to the initialization month.

Another critical gap is the lack of systematic comparison of bias correction and downscaling methods—such as BCSD, MOS, and SCA—when applied to CanSIPS outputs over Iranian stations. While these methods have been widely employed in North America, their relative performance in Iran's diverse climatic settings remains unexplored. To address these gaps, the present study conducts a dedicated, station-based, and disaggregated evaluation of CanSIPS forecasts for Tmin, Tmax, and precipitation across 30 Iranian synoptic stations. We assess forecast skill across distinct climate regimes, analyze its dependence on lead time and initialization month, and comparatively evaluate the effectiveness of multiple bias correction techniques. The findings are expected to provide a robust scientific basis for operationalizing seasonal forecasts in Iran, supporting informed decision-making in agriculture, water resources, and climate adaptation strategies.