National Weather Service United States Department of Commerce

STI Modeling Program Seminars

STI has initiated a seminar series to facilitate communication among researchers and model developers. Slides will be posted and available for download before the presentation.

Most recent Seminar: Monday, September 25, 2017   

Tim Palmer, Royal Society Research Professor, University of Oxford



More accuracy with less precision - assessing information content for reliable weather and climate prediction

Presentation slides



The reasons for stochastic parametrisation are reviewed. From this, we conclude that conventional 64-bit floating-point representations in NWP and climate models are, for the most part, unnecessary and energetically profligate. It is argued that a new start is needed for the development of Earth-System Models, based on the primacy of information content.

Previous Seminars

The stratiform region of squall lines and its representation in convection-allowing numerical models 

By George H. Bryan, (NCAR ) presented June 14, 2017

Presentation slides


The stratiform region of a squall line is the mesoscale area of relatively low radar reflectivity (20-40 dBZ) that often develops near the more-intense convective region (40-60 dBZ). Real-time numerical models that explicitly represent convection (with grid spacing of ~3 km) often fail to produce stratiform regions. This talk focuses on some microphysical and dynamical processes that can improve the structure of the stratiform region in such models. A key microphysical process is the inclusion of a fast-falling frozen hydrometeor (i.e., hail), which is primarily associated with the high-reflectivity convective region, but allows for the production of slower-falling snow aloft which ultimately forms the stratiform region. Turbulent mixing processes also play a key role, as revealed by high-resolution simulations (with grid spacing ~100 m) that show how air in stratiform regions has a history of mixing between mid-level and boundary-layer air. Identical simulations with relatively coarser resolution (grid spacing ~1 km) show practically no mixing and, consequently, inaccurate humidity profiles. In summary, stratiform regions are produced by a combination of microphysical and dynamical processes, and thus efforts to improve convection-resolving models need to address both processes.

A joint MAPP-NGGPS Webinar as part of the OneNOAA Seminar Series: Sources of Predictability at Subseasonal to Seasonal Time Scales


Speakers: Frédéric Vitart (European Centre for Medium-Range Weather Forecasts (ECMWF)), Cristiana Stan (George Mason University), Antje Weisheimer (University of Oxford & ECMWF)


Presented May 24, 2107 

Sponsors: NOAA OAR/CPO Modeling, Analysis, Predictions, and Projections program and NOAA NWS Next Generation Global Prediction System program

Seminar POC:

  • MJO Prediction and Teleconnections in Sub-seasonal Forecasts(Frédéric Vitart)


    The Madden Julian Oscillation (MJO) is the dominant intra-seasonal mode of organized convective activity in the Tropics, with also a considerable impact in the middle and high latitudes. The skill of sub-seasonal forecasting systems to predict the MJO has improved significantly over the past decade, although most models still have difficulties propagating the MJO across the Maritime continent. The MJO predictive skill and teleconnections in the high latitudes have been diagnosed in 10 operational sub-seasonal prediction models from the WWRP/WCRP Sub-seasonal to Seasonal Prediction (S2S) database. Results suggest that the S2S models display skill to predict the MJO between 2 and 4 weeks, although the majority of S2S models tend to produce a too weak and slow propagating MJO in the extended forecast range. All the S2S models produce MJO extratropical teleconnections which are too weak over the Euro-Atlantic sector, which suggests that they do not fully exploit the predictability associated to the MJO in the Northern Extratropics. The impact of model resolution and ocean-atmosphere coupling on the MJO prediction skill and teleconnections will be discussed.


    The subseasonal-to-seasonal variability of Northern Hemisphere midlatitudes and its influence on forecasts for weeks 3-4 (Cristiana Stan) 

    Abstract: The variability of the extra-tropics and its interaction with the tropics is studied at seasonal and intraseasonal time scales. Nonlinear oscillations in the extra-tropics are extracted from daily anomalies of 500-hPa geopotential for the period 1979-2012 using a data-adaptive method. One of the emerging global oscillations has a period of 120 days and over the North Atlantic region its pattern resembles the canonical North Atlantic Oscillation (NAO). A composite of the lifecycle of the 120-day oscillation shows that for the phase peak (phase 2) the pattern consists of a surface pressure dipole over the North Atlantic accompanied by cold (warm) surface temperature anomalies and anticyclonic wind at low-level to the north (south).

    The 120-day oscillation was included as a predictor in an experimental version of the statistical forecast model used at CPC as part of the forecast tools for the Week 3-4 temperature and precipitation outlook. The influence of the predictor leads to mixed improvement and deterioration in cross-validated skill. The impacts on Week 3-4 U.S. temperature prediction appear most complimentary in boreal winter and when the MJO is over the Maritime Continent or West Pacific. Less skill is generally added when the MJO is over the Indian Ocean or forecast occurs during boreal summer.


    Atmospheric Seasonal Forecasts of the 20th Century: Multi-Decadal Variability in Predictive Skill of the Winter NAO (Antje Weisheimer)

    Abstract: Based on skill estimates from hindcasts made over the last couple of decades, recent studies have suggested that considerable progress has been made in forecasting winter climate anomalies over the Euro-Atlantic area using current-generation forecast models. However, previous-generation models had already shown that forecasts of winter climate anomalies in the 1960s and 1970s were less successful than forecasts of the 1980s and 1990s. Given that the more recent decades have been dominated by the NAO in its positive phase, it is important to know whether the performance of current models would be similarly skilful when tested over periods of a predominantly negative NAO.

    To this end, a new ensemble of retrospective atmospheric seasonal forecasts covering the period 1900 to 2009 has been created, providing a unique tool to explore many aspects of atmospheric seasonal climate prediction. In this study we focus on the multi-decadal variability in predicting the winter NAO. The existence of relatively low skill levels during the period 1950s -1970s has been confirmed in the new dataset. This skill appears to increase again for earlier and later periods. Whilst these interdecadal differences in skill are, by themselves, only marginally statistically significant, the variations in skill strongly co-vary with statistics of the general circulation itself suggesting that such differences are indeed physically real. The mid-Century period of low forecast skill coincides with a negative NAO phase but the relationship between the NAO phase/amplitude and forecast skill is more complex than linear.


    Subscribe to the OneNOAA Science Seminar weekly email: Send an email to with the word `subscribe' in the subject or body. See

Winter forecast skill: teleconnections from the tropics and a case study for winter 2015/16

By Adam Scaife (Met Office) presented March 8, 2017


Presentation slides


Part 1: Skillful climate predictions of the winter North Atlantic Oscillation and Arctic Oscillation out to a few months ahead have recently been demonstrated, but the source of this predictability remains largely unknown. Here we investigate the role of the tropics in this predictability. We show high levels of skill in tropical rainfall predictions, particularly over the Pacific but also the Indian and Atlantic Ocean. Rainfall fluctuations in these regions are associated with clear signatures in the atmospheric circulation that are approximately symmetric about the equator in boreal winter. We show how these patterns can be explained as steady poleward propagating linear Rossby waves emanating from just a few key source regions and that predicted tropical rainfall explains a significant fraction of the predicted variation of the winter North Atlantic Oscillation.

Part 2: The northern winter of 2015/16 gave rise to the strongest El Niño event since 1997/8. Central and eastern Pacific sea surface temperature anomalies exceeded three degrees and closely resembled the strong El Niño in winter of 1982/3. A second feature of this winter was a strong westerly phase of the Quasi-Biennial Oscillation and very strong winds in the stratospheric polar night jet. At the surface, intense extratropical circulation anomalies occurred in both the North Pacific and North Atlantic that were consistent with known teleconnections to the observed phases of ENSO and the QBO. The North Atlantic Oscillation was very positive in the early winter period (Nov-Dec) but the flow was blocked in late winter. Initialised climate predictions were able to capture these signals at seasonal lead times. This case study adds to the evidence that north Atlantic circulation exhibits predictability on seasonal timescales, and we show that even aspects of the sub-seasonal evolution were predicted in this case.

Sub-seasonal prediction of aerosols fields and impact on meteorology using the ECMWF's coupled Ensemble Prediction System

By Angela Benedetti and Fréderic Vitart (ECMWF) presented February 8, 2016


Presentation Slides

Abstract: Recent years have seen the rise of global operational atmospheric composition models for several applications including climate monitoring, provision of boundary conditions for regional air quality forecasting, and energy sector applications, to mention a few. Typically global forecasts are provided in the medium-range up to five days ahead. In this work we investigate the feasibility of sub-seasonal to seasonal prediction of aerosols using the ECMWF’s coupled Ensemble Prediction System. The motivation of this study is also to investigate the impact of the aerosol direct effect on meteorological variables such as winds, temperature and precipitation. A comparison between a run with fully prognostic and interactive aerosols and a control run using the current operational set-up will be presented. Aerosol forecast fields at the weekly/monthly scales will also be presented and compared with corresponding analysis to assess their quality.