Systematic Error Analysis and Calibration of 2-m Temperature for the NCEP GEFS Reforecast of the Subseasonal Experiment (SubX) Project

AbstractThe National Centers for Environmental Prediction has generated an 18-yr (1999-2016) subseasonal (weeks 3&4) reforecast to support the Climate Prediction Center’s operational mission. To create this reforecast, the subseasonal experiment version of the GEFS was run every Wednesday, initialized at 0000 UTC with 11 members. The Climate Forecast System Reanalysis (CFSR) and Global Data Assimilation System (GDAS) served as the initial analyses for 1999-2010 and 2011-2016, respectively. The analysis of 2-m temperature error demonstrates that the model has a strong warm bias over the Northern Hemisphere (NH) and North America (NA) during the warm season. During the boreal winter, the 2-m temperature errors over NA exhibit large inter-annual and intra-seasonal variability. For NA and the NH, weeks 3&4 errors are mostly saturated, with initial conditions having a negligible impact. Week 2 errors (day-11) are ~88.6% and 86.6% of their saturated levels, respectively.The 1999–2015 reforecast biases were used...