Abstract

A high-resolution global model forecast of the formation of Hurricane Frederic of 1979 is analyzed by means of several diagnostic computations on the model's output history. The formation is addressed from an analysis of limited-area energetics where the growth of eddy kinetic energy is examined. The question on internal versus external forcing during the formative stage of the hurricane is explored by means of the Kuo-Eliassen framework for the radial-vertical circulation of the hurricane. The intensity of the predicted hurricane is diagnosed from a detailed angular momentum budget following the three-dimensional motion of parcels arriving at the maximum wind belt. Overall, the successful simulation of the hurricane has enabled us to make such a detailed diagnosis of the predicted hurricane at a high resolution. The principal findings of this study are that a north-south-oriented beating function maintained a zonal easterly flow that supplied energy barotropically during the growth of an African wave. The growth of eddy kinetic energy is somewhat monotonic and slow throughout the history of the computations. The initial development of the easterly wave appears to be related to the widespread weak convective heating that contributes to a covariance of heating and temperature and of temperature and vertical velocity. The hurricane development period is seen as one where both the barotropic and convective processes contribute to the growth of eddy kinetic energy. During this developing stage, the growth of radial-vertical circulation is largely attributed to convective, radiative, and frictional forcings. The role of eddy convergence of momentum flux appears to be insignificant. The intensity issue of the storm (maximum wind of the order of 37 m s⁻¹) was addressed by means of a detailed angular momentum budget following parcel motion. The pressure torque in the model simulation had a primary role in explaining the intensity of the predicted storm. It is only in the storm's inner rain area where the frictional stress becomes quite large. But at these small radii the frictional torque is still smaller compared to the contribution from the (small but significant) azimuthal asymmetries of the pressure field and the resulting pressure torques.

Abstract

In this paper the authors illustrate wave trains that are excited during the equatorial passage of the annual cycle of monsoonal convection. Twice a year, during roughly the months of December–January and March–April, the annual cycle of monsoonal convection crosses the equator. A principle axis of annual cycle monsoon precipitation extends from the Java Sea to the eastern Himalayas. Monsoonal convection makes a north–south seesaw roughly along this axis each year. Near-equatorial convection provides a tropospheric heat source somewhat akin to that of El Niño over the equatorial Pacific Ocean. This equatorial passage of the monsoonal heat source excites a wave train, somewhat similar to the familiar Pacific–North American pattern. Monsoonal wave trains were extracted from a 9-yr dataset, and a composite geometry was constructed. This note also illustrates excitation of short-period wet and dry spells associated with excitation of this wave train. This is illustrated for several trough and ridge locations of the wave train by examining rainfall for a sequence of days some 10 days prior to and 10 days subsequent to passage of this wave train. There is a strong suggestion that equational passage of monsoon convection does influence short-term dry and wet spells along the wave train; that is, beneath upper troughs (ridges), wet (dry) weather prevails.

Abstract

This paper presents some recent results on physical initialization from the use of a very high resolution global model. Fundamentally this procedure improves the model-based initial rainfall, surface fluxes, and diagnostic cloud amount. Physical initialization is a useful procedure for the nowcasting of rainfall. Correlation between model-based initialized rain and satellite/rain gauge-based rain over the Tropics (for 6-h averages and averaged over transform grid squares) is of the order 0.85. This compares with a correlation of around 0.3 for models that do not include physical initialization. The day 1 tropical rainfall forecast skill is also relatively high for the physically initialized experiments; the correlation is of the order 0.55. It should be noted that the lifetime of mesoconvective systems is approximately 1 day, whereas more organized tropical disturbances may last substantially longer. A major portion of the tropical rainfall is associated with these short-lived systems, hence the skill beyond 1 day degrades somewhat. However, the model does seem to capture the 1-day passage of mesoconvective systems and their coupling to the large-scale, synoptic environment. The mesoconvective systems illustrated exhibit a robust vertical structure of divergence, heating, and vertical motion, which is absent without physical initialization.

The organization of mesoconvective systems (advected by the large-scale circulations and coalescence of the mesoscale elements) appears to play an important role in the formation of tropical storms. The vorticity associated with these mesoscale elements, however, does not exhibit any interesting organization during the forecast as the storms form. The Florida State University atmospheric global circulation model at the resolution T213 discerns the tight central circulation features and the outer rainbands of Hurricane Andrew (1992), which appear similar to the radar imagery; however, the storm as seen from the model is not on the exact scale as that of the radar that is shown. Further enhancement of resolution is needed to model tropical storms on a more realistic scale, which is well known in the modeling community. Overall this study demonstrates that mesoconvective elements are in fact simulated by very high resolution global models. It appears that very high resolution models with an augmented analysis using satellite data may soon aid in resolving the formation issue associated with tropical cyclones and cyclogenesis.

Abstract

Ensemble forecasting of hurricane tracks is an emerging area in numerical weather prediction. In this paper, the spread of the ensemble of forecast tracks from a family of different First Global GARP (Global Atmospheric Research Program) Experiment analyses is illustrated. All forecasts start at the same date and use the same global prediction model. The authors have examined ensemble forecasts for three different hurricanes/typhoons of the year 1979. The authors have used eight different initial analyses to examine the spread of ensemble forecasts through 6 days from the initial state. A total of 16 forecasts were made, of which 8 of them invoked physical initialization. Physical initialization is a procedure for improving the initial rainfall rates consistent with satellite/rain gauge based measures of rainfall. The main results of this study are that useful track forecasts are obtained from physical initialization, which is shown to suppress the spread of the ensemble of track forecasts. The spread of the tracks is quite large if the rain rates are not initialized. The major issue here is how one could make use of this information on ensemble forecasts for providing guidance. Toward that end, a statistical framework that makes use of the spread of forecast tracks to provide such guidance is presented.

Abstract

This study is based on a global coupled atmosphere–ocean model climate prediction that was designed to include 14 layers over the atmosphere and 17 layers within the ocean. In this model an 11-yr data assimilation includes physical initialization of the daily rainfall estimates. No flux corrections are included in the seasonal and annual forecasts of this coupled model. It is first shown that intraseasonal oscillation on the Madden–Julian timescale was an important feature during the onset of the El Niño of 1997. It is shown that this feature is retained in the model’s data assimilation and in the forecasts. The forecasts commence on 1 April 1997. The model forecasts showed an El Niño warming of the equatorial Pacific Ocean waters commencing with the excitation of a Kelvin wave. The Niño-3.4 region acquired above-normal sea surface temperature anomalies (SSTAs) by 15 May. The warm SSTs reached a peak by around January 1998. The El Niño made its demise by June 1998. The life cycle of the entire SSTA shows remarkable agreement to the observed anomalies over the Pacific Ocean. The subsurface temperature anomalies exhibit eastward propagating subsurface warm and cold water that are in phase with the El Niño and the La Niña features at the surface. Phenomenologically, this study is quite successful in showing the following.

• Velocity potential anomalies at the 200-hPa level are good indicators for long-lasting dry spells. In particular the authors have remarkable success in predicting the long-lasting dry spell over Florida (which resulted in major fires over Florida during June 1998, some 14 months into the forecast) and over Indonesia (which resulted in major fires over Indonesia during September and October 1997). This was by far the most promising result of the coupled modeling study. This study also enumerates several areas of the climate of 1997–98 that were not reasonably simulated at the present resolution of the coupled model. The model does not exhibit very high skill in prediction of precipitation anomalies over the Asian–Australian monsoon world, which is most likely due to the resolution and organization of convection issues.

• A realistic picture is shown of the North American monsoon system (the Mexico–Arizona monsoon) with wet conditions along 110°W, dry conditions along 95°W, and wet conditions along 80°W during the summers of 1997 and 1998. Furthermore, the model successfully shows a stronger North American monsoon system during the post–El Niño year 1998 compared to the El Niño year 1997. This is in accordance with the climatological and observational findings.

• California rainfall during January and February 1998, arising from the eastward passage of disturbances from the Pacific Ocean, was successfully simulated, although the rainfall amounts at the model resolution were roughly one-third of the observed peak estimates.