Assimilation of GOES-16 Radiances and Retrievals into the Warn-on-Forecast System

Abstract

The increasing maturity of the Warn-on-Forecast System (WoFS) coupled with the now operational GOES-16 satellite allows for the first time a comprehensive analysis of the relative impacts of assimilating GOES-16 all-sky 6.2-, 6.9-, and 7.3-μm channel radiances compared to other radar and satellite observations. The WoFS relies on cloud property retrievals such as cloud water path, which have been proven to increase forecast skill compared to only assimilating radar data and other conventional observations. The impacts of assimilating clear-sky radiances have also been explored and shown to provide useful information on midtropospheric moisture content in the near-storm environment. Assimilation of all-sky radiances adds a layer of complexity and is tested to determine its effectiveness across four events occurring in the spring and summer of 2019. Qualitative and object-based verification of severe weather and the near-storm environment are used to assess the impact of assimilating all-sky radiances compared to the current model configuration. We focus our study through the entire WoFS analysis and forecasting cycle (1900–0600 UTC, daily) so that the impacts throughout the evolution of convection from initiation to large upscale growth can be assessed. Overall, assimilating satellite data improves forecasts relative to radar-only assimilation experiments. The retrieval method with clear-sky radiances performs best overall, but assimilating all-sky radiances does have very positive impacts in certain conditions. In particular, all-sky radiance assimilation improved convective initiation forecast of severe storms in several instances. This work represents an initial attempt at assimilating all-sky radiances into the WoFS and additional research is ongoing to further improve forecast skill.

Evaluation of an experimental Warn‐on‐Forecast 3DVAR analysis and forecast system on quasi‐real‐time short‐term forecasts of high‐impact weather events

“…The poor performance in Kansas in the 17 May case is likely due to a warm and dry bias of the near‐storm environment. As discussed above, assimilating more observations such as boundary‐layer observations and/or satellite radiance (such as Jones et al ., 2020) may help alleviate this problem.…”

“…Seeking to assimilate more available observations, especially those depicting the near-storm environment, could potentially help alleviate this drawback. For example, using the GSI-EnKF DA system, Jones et al (2020) demonstrated that a favorable (moister and colder) environment is produced by assimilating GOES 16 all-sky radiances in addition to radar observations (their ALL experiment, Figures 6d and 7b), and the simulated dryline position is near the location of the observed convection initiation. Therefore, as seen from their figures 4d and 5d, only forecasts with the additional favorable environment can successfully predict the initiation and subsequent development of the Kansas storm, though with slightly larger coverage compared with the observations.…”

Evaluating the Impact of Planetary Boundary Layer, Land Surface Model, and Microphysics Parameterization Schemes on Cold Cloud Objects in Simulated GOES‐16 Brightness Temperatures

“…Object-based methods present a powerful way to analyze cloud cover forecast accuracy because traditional metrics typically penalize CAM forecasts for displacement errors between the forecast and observed cloud objects. Therefore, object-based methods and IR BTs have been used to assess forecast accuracy (Griffin et al, 2017a(Griffin et al, , 2017b(Griffin et al, , 2020Jones et al, 2018Jones et al, , 2020Senf et al, 2018;Skinner et al, 2016Skinner et al, , 2018. In this analysis, cloud objects are identified in simulated and observed satellite imagery using the Method for Object-Based Diagnostic Evaluation (MODE; Bullock et al, 2016;Davis et al, 2006aDavis et al, , 2006bDavis et al, , 2009.…”

Correlation Structures between Satellite All-Sky Infrared Brightness Temperatures and the Atmospheric State at Storm Scales

“…Ensemble-based data assimilation techniques, such as the ensemble Kalman filter (EnKF), play an important role in recent advances in the assimilation of all-sky infrared (IR) brightness temperatures (BTs) into regional numerical weather prediction (NWP) models. Different variations of the EnKF are used in numerous observing system simulation experiments (OSSEs; e.g., Otkin, 2010Otkin, , 2012Zupanski et al, 2011;Jones et al, 2013Jones et al, , 2014Cintineo et al, 2016;Zhang et al, 2016a;Minamide and Zhang, 2017) and realdata studies (e.g., Zhang et al, 2016aZhang et al, , 2018Zhang et al, , 2019aZhang et al, , b, 2021Honda et al, 2018a, b;Minamide and Zhang, 2018;Okamoto et al, 2019;Otkin and Potthast, 2019;Chan et al, 2020;Jones et al, 2020). The EnKF uses flow-dependent background error covariances, and several studies show that the EnKF is more suitable for convective phenomena at the mesoscales and the storm scales than traditional variational techniques (e.g., Meng and Zhang, 2008;Zhang and Zhang, 2012;Schwartz and Liu, 2014;Johnson et al, 2015).…”