In cold region, conceptual models assigned precipitation phase, liquid (rain) or solid (snow), cause vastly different atmospheric, hydrological, and ecological responses, along with significant differences in evaporation, runoff, and infiltration fates for measured precipitation mass. A set air temperature threshold (ATT) applied to the over 30% annual precipitation events occurring with surface air temperatures between −3 and 5 °C resulted in 11.0 and 9.8% misclassified precipitation in Norway and Sweden, respectively. Surface air temperatures do not account for atmospheric properties causing precipitation phase changes as snow falls toward the ground. However, cloud base height and relative humidity (RH) measured from the surface can adjust ATT for expected hydrometeor-atmosphere interactions. Applying calibrated cloud base height ATTs or a linear RH function for Norway (Sweden) reduced misclassified precipitation by 4.3% (2.8%) and 14.6% (8.9%) misclassified precipitation, respectively. Cloud base height ATTs had lower miss-rates with low cloud bases, 100 m in Norway and 300 m in Sweden. Combining the RH method with cloud base ATT in low cloud conditions resulted in 16.1 and 10.8% reduction in misclassified precipitation in Norway and Sweden, respectively. Therefore, the conceptual model output should improve through the addition of available surface data without coupling to an atmospheric model.

在寒冷地区，分配了降水阶段、液体（雨）或固体（雪）的概念模型会导致大不相同的大气、水文和生态响应，以及测量降水质量的蒸发、径流和入渗命运的显着差异。将设定的气温阈值 (ATT) 应用于地表气温在 -3 至 5 °C 之间的年降水量超过 30% 的事件，导致挪威和瑞典的降水误分类分别为 11.0% 和 9.8%。地表气温不考虑大气特性，当雪落向地面时会导致降水相变化。然而，从地表测量的云底高度和相对湿度 (RH) 可以调整 ATT 以适应预期的水汽-大气相互作用。对挪威（瑞典）应用校准的云底高度 ATT 或线性 RH 函数分别将误分类降水减少了 4.3% (2.8%) 和 14.6% (8.9%) 误分类降水。云底高度 ATT 的误报率较低，云底较低，挪威为 100 m，瑞典为 300 m。在低云条件下将 RH 方法与云底 ATT 相结合，分别使挪威和瑞典的误分类降水减少了 16.1% 和 10.8%。因此，概念模型输出应该通过添加可用的表面数据而不与大气模型耦合来改进。挪威 100 m，瑞典 300 m。在低云条件下将 RH 方法与云底 ATT 相结合，分别使挪威和瑞典的误分类降水减少了 16.1% 和 10.8%。因此，概念模型输出应该通过添加可用的表面数据而不与大气模型耦合来改进。挪威 100 m，瑞典 300 m。在低云条件下将 RH 方法与云底 ATT 相结合，挪威和瑞典的误分类降水分别减少了 16.1% 和 10.8%。因此，概念模型输出应该通过添加可用的表面数据而不与大气模型耦合来改进。