Based on the fifth generation global reanalysis data (ERA5) of the European Centre for Medium-Range Weather Forecasts (ECMWF), the model climate is constructed. The ensemble mean abnormal temperature forecast and abnormal temperature probability forecast products for China are constructed according to the maximum and minimum temperatures by using ECMWF ensemble forecast and standardized anomaly forecast method. The forecast performance was tested and compared with the extreme forecast index (EFI). Furthermore, based on “abnormal temperature impact matrix”, forecast index of impact degree of abnormal temperature event was constructed. The forecast application of related products was discussed through the cases of abnormal temperature events. The results show that the ensemble mean abnormal temperature forecast product based on ERA5 and ECWMF ensemble forecast has a good forecasting effect for both summer and winter abnormal temperature events in China, and the forecast performance is better than or close to EFI through comparative verification. The product can be used as a supporting product for forecasting abnormal temperature events. Abnormal temperature probability forecast products can reflect the information of abnormal temperature events predicted in the ensemble members of the set, have advantages in finding early abnormal temperature event signals in medium-range forecast, and can reflect the uncertainty information of abnormal weather forecast. The forecast index of impact degree of abnormal temperature event combines the prediction information of the probability of abnormal weather and the prediction information of the anomaly of abnormal weather. It can give the objective and quantitative prediction results of the abnormal temperature time in one product. The prediction of the abnormal low temperature event caused by a cold surge weather process proves that this index has a relatively good prediction effect. It has certain indicative significance and application prospect for the prediction and early warning of abnormal temperature events.