Sub-Seasonal Precipitation and Surface Temperature Forecast Skill Using the Extended Range NCEP Global Ensemble Forecasting System
Abstract
Efforts at NOAA for seamless prediction across spatial and temporal scales have put focus on improving sub-seasonal (weeks 3 and 4) forecasts. The next implementation of the NCEP Global Ensemble Forecasting System (GEFS) will extend select forecast cycles from 16 day to 35 day forecasts. This will provide sub-seasonal ensemble forecasts with increased spatial resolution and larger membership compared with the current Climate Forecasting System version 2 (CFSv2) that covers this time scale. The nonlinear evolution of the atmosphere places a predictability limit on specific weather events, but anomalous weekly and longer-term averages can provide forecast skill in the sub-seasonal time scale due to slowly evolving forcing (e.g. sea surface temperature, soil moisture, and sea ice) and low-frequency variability (e.g. El Nino, MJO, and PNA) retained and propagated from the initial atmospheric state.The current NCEP GEFS is an uncoupled atmospheric model with prescribed sea surface temperatures (SST). Using an extended range configuration of the GEFS, the impact of various SST forcing on week 3 and 4 forecast skill of accumulated precipitation and surface temperature are examined across North America. The experiment SST configurations include a 90-day e-folding SST towards observed climatology (control), control with imposed Near Sea Surface Temperature (NSST) diurnal variations, observed SSTs, and raw and bias corrected SSTs from the CFSv2 (two-tiered). A tercile Heidke Skill Score (HSS) is used to assess performance of the GEFS SST experiments as well as the operational CFSv2, all validated against GEFS analyses and observations. Preliminary results indicate little to no skill for accumulated precipitation with some skill for surface temperatures. The two-tiered experiments as well as observed SSTs improve surface temperature HSS compared to control. All GEFS experiments have larger surface temperature HSS at all lead times compared with CFSv2. The surface temperature results are encouraging given the extended GEFS is tasked to provide an additional source of information, complementary to the CFSv2, to aid the Climate Prediction Center in generating week 3 and 4 outlooks.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFM.A33J0404M
- Keywords:
-
- 3337 Global climate models;
- ATMOSPHERIC PROCESSESDE: 0550 Model verification and validation;
- COMPUTATIONAL GEOPHYSICSDE: 1817 Extreme events;
- HYDROLOGYDE: 4341 Early warning systems;
- NATURAL HAZARDS