The type of land surface model used in climate simulations can significantly impact the accuracy of extreme temperature and precipitation predictions. A study using the WRF model over North America found that different land surface models led to varying results in the intensity and duration of extreme weather events. For example, heat waves could be up to 5 degrees Celsius warmer and last 6 days longer depending on the land surface model used. This highlights the importance of selecting the right land surface model to reduce uncertainties in climate model projections and improve the accuracy of extreme weather predictions.