This study aims to understand the fluctuations observed in Hurricane Irma (2017), which change the tangential wind speed and the size of the radius of maximum surface wind and therefore affect short-term destructive potential. Intensity fluctuations observed during a period of rapid intensification of Hurricane Irma between 4 and 6 September 2017 are investigated in a detailed modelling study using an ensemble of Met Office Unified Model (MetUM) convection-permitting forecasts. Although weakening and strengthening phases were defined using 10 m wind, structural changes in the storm were observed through the lower troposphere, with the most substantial changes just above the boundary layer (at around 1500 m). Isolated regions of rotating deep convection, coupled with outward propagating vortex Rossby waves, develop during the strengthening phases. Although these isolated convective structures initially contribute to the increase in azimuthally averaged tangential wind through positive radial eddy vorticity fluxes, the continued outward expansion of convection eventually leads to a negative radial eddy vorticity flux, which halts the strengthening of the tangential wind above the boundary layer at the start of the weakening phase. The outward expansion of the azimuthally averaged convection also enhances the outflow above the boundary layer in the eyewall region, as the convection is no longer strong enough to ventilate the mass inflow from the boundary layer in a process similar to one described in a recent idealised study.
One of the biggest challenges in weather forecasting is predicting when a tropical cyclone (TC) will rapidly intensify. Rapid intensification is defined as a rate of surface wind increase of at least 15.4 m s
The simplest paradigm for TC intensification can be understood by considering the case of a stationary vortex in gradient wind balance.
The boundary layer spin-up mechanism, as described by
The axisymmetric theory does not fully explain the development of a TC, particularly during rapid intensification, due to the presence of asymmetric processes. These include the role of isolated regions of deep rotating convection, which are local small regions of high relative vorticity and high vertical velocity within the eyewall. Isolated regions of deep rotating convection and their associated downdrafts can act to transport heat and angular momentum inwards to the eye prior to rapid intensification
Many of these unbalanced and asymmetric processes have been examined in studies of intensity fluctuations that occur during the intensification of TCs, which are not easily explained by an axisymmetric balanced dynamical theory. One example is vacillation cycles, a form of intensity fluctuations that sometimes occurs during rapid intensification.
Another form of intensity fluctuation was identified recently in
Hurricane Irma (2017) underwent rapid intensification twice (Fig.
In this paper we analyse the intensity fluctuations of Hurricane Irma using both observations and convection-permitting ensemble simulations to help to understand whether or not the inner core intensity fluctuations are a previously unknown phenomenon or exist on a spectrum that may include vacillation cycles, eyewall replacement cycles, or other structural changes that occur during rapid intensification. The analysis will involve investigating the cause of the intensity fluctuations and understanding the structural and dynamical changes of the TC in the transition between a strengthening and weakening phase.
The paper will be organised in the following way: Sect.
Hurricane Irma was the first major hurricane of the 2017 North Atlantic hurricane season. Irma peaked at an intensity of 80 m s
Irma formed out of an African easterly wave off the west coast of Africa at around 30
The second period of rapid intensification began on 04 September, with Irma intensifying from a Category 3 storm (945 hPa, 50 m s
Despite favourable environmental conditions, with a low vertical wind shear, high sea surface temperatures, and adequate mid-level moisture, Irma weakened to Category 4 during 7 September due to the start of an eyewall replacement cycle. Irma passed over Little Inagua at 05:00 UTC on the same day.
Thereafter, apart from a brief period of intensification that occurred around 03:00 UTC on 9 September, Irma gradually weakened due to increasing vertical wind shear and eventually land interaction after making landfall in Florida on 11 September. Irma finally dissipated inland on 13 September. Further details on the synoptic overview of Hurricane Irma (2017) are available in
A key source of observational data were aircraft flyovers. Multiple flights were made through Hurricane Irma operated by the National Oceanic and Atmospheric Administration (NOAA). The flyovers were conducted with aircraft from the NOAA aircraft operations centre and the 53rd Weather Reconnaissance Squadron. Observations used from these flights include in situ wind speed and pressure measurements, dropsondes, and airborne radar. Satellite-visible, infra-red (IR), and morphed integrated microwave imagery
The SATCON intensity estimates are derived from the structure of the TC with heavy usage of microwave and satellite IR imagery, so relating structural changes to intensity changes would be a circular argument. Where possible, therefore, mean sea level pressure (MSLP) data from flights and dropsondes are also used for short periods where there are a large number of flyovers such as in the afternoon of 5 September. MSLP data are preferable to tangential wind data as an intensity proxy, because the latter are strongly dependent on the direction of the flight into the eyewall and the height of the aircraft.
The dropsonde data are available in a quality-controlled post processed format (in some cases raw data were used instead due to lack of availability). In addition, some of the NOAA aircrafts are equipped with C-band and Doppler radars on the nose, lower fuselage, and tail. The processed lower fuselage and tail radar data are used in the analysis and show precipitation in dBZ reflectivity. All the processed dropsonde and flight-level data used in this analysis are available from the Hurricane Research Division (
The focus of the analysis is on the second period of rapid intensification which starts on 4 September at around 00:00 UTC and finishes around 00:00 UTC on 6 September (Figs.
Observed minimum sea level pressure as a function of time based on SATCON and National Hurricane Center (NHC) forecaster-assessed best track estimates as well as direct dropsonde and flight measurements. The 96 h period shown is the same as the simulation initialised on 3 September 00:00 UTC. Two notable weakening or stagnation periods during the period of rapid intensification are highlighted by the blue bands.
Figure
NOAA P3 flight-level radar (in dBZ) on
An 18-member ensemble of convection-permitting forecasts for Hurricane Irma has been produced using a limited-area configuration of the Met Office Unified Model
The MetUM solves the fully compressible, deep-atmosphere, non-hydrostatic equations of motion using a semi-implicit, semi-Lagrangian scheme (see
Both the MetUM and JULES include a comprehensive set of parameterisation schemes for key physical processes, and the way in which these are configured defines a model science configuration. Here we use the tropical version of the Regional Atmosphere and Land – Version 1 (RAL1-T) configuration presented in
The extent of the regional model domain is shown in Fig.
Each member of the convection-permitting ensemble is one-way nested inside a corresponding member of the Met Office global ensemble prediction system, MOGREPS-G
MOGREPS-G includes two stochastic physics schemes to represent the effects of structural and subgrid-scale model uncertainties: the random parameters scheme
Incorporated into the MetUM are two sets of tracers (PV and potential temperature,
Much of the analysis is done from an axisymmetric perspective in storm-relative cylindrical coordinates. Calculations such as this can be highly sensitive to the location of the storm centre. The simplest way to find the TC centre in the model simulation is to find the coordinates that minimise the surface level pressure field. However, mesovortices within the eyewall often lead to the minimum surface level pressure being displaced from the geometric centre of the eye into the inner eyewall, which can cause the tangential flow within the eye to be overestimated and the tangential flow outside the eye to be underestimated. Several more robust methods have been proposed, each with their own advantages and disadvantages. These include finding PV centroids
The method used in this analysis balances the need for a consistent and reliable method for finding the location of the TC centre to an appropriate degree of precision, while considering the computational cost of doing so for 18 ensemble members over a 4 d simulation period. The method used here is similar to the one used by
The convergence criteria for the algorithm are the following: no more than 50 function evaluations, an absolute error between iterations of no more than 0.5 m s
The fluctuations modelled during rapid intensification in Hurricane Irma have similarities to both vacillation cycles and eyewall replacement cycles but with important differences that will be discussed in detail.
The second period of rapid intensification in Hurricane Irma is broadly captured by the convection-permitting ensemble forecasts (Fig.
Various model diagnostics (solid lines) and corresponding observations (dashed lines, where available) as a function of time. Details are given in the legend. Blue bands indicate weakening phases, and red bands indicate strengthening phases during the rapid intensification period. The individual strengthening and weakening phases have been labelled (see top of plot). W stands for “weakening”, S stands for “strengthening”. Phases have been subjectively identified. The RMW refers to the surface or 1532 m radius of maximum azimuthally averaged tangential wind speed. The tangential wind is the azimuthally averaged tangential wind at the RMW.
By examining the change in the 10 m total wind speed, MSLP, and 10 m RMW over time (Fig.
The maximum tangential wind, particularly near the top or just above the boundary layer (e.g. at 1532 m), also exhibits these fluctuations but does lag behind compared to higher levels (e.g. at 3002 m), where the maximum tangential wind follows a similar pattern to the 10 m total wind speed. The lag is also present in the expansion of the RMW, with the increase in the RMW happening at 1532 m (dark green line) prior to the increase in the 10 m RMW (aqua line). At the surface, the signal in the tangential wind speed is weaker compared to at higher levels. The role the radial flow plays in modifying the total surface wind speed during the fluctuations and the reason for the tangential wind spin-down preceding a weakening phase is explored in detail in Sect.
The simulation shows four weakening periods and three strengthening periods which are defined in terms of 10 m total wind speed, 10 m RMW, and MSLP. There is also an uninterrupted period of intensification prior to these fluctuations. During the period of intensity fluctuations from 45 to 84 h, Irma is still rapidly intensifying overall, so the brief interruptions in intensification do not stop rapid intensification from happening. The main aim of the analysis is to understand better why these intensity fluctuations happen during this period of rapid intensification and to gain insight into elements of the mechanism behind the fluctuations and any structural changes with which they are associated. Although the intensity fluctuations have been defined with respect to the surface, for the purposes of easy comparison with observational data, the most dramatic changes occur just above the boundary layer, so the subsequent analysis will focus on the 1500 m level and how structural changes at this level impact the boundary layer below it.
It should be noted that during the analysed rapid intensification period, Hurricane Irma was a fairly symmetric storm under low vertical wind shear, with environmental factors playing a minimal role in these fluctuations. Changes in vertical shear, translation speed, sea surface temperature, maximum potential intensity, and the diurnal cycle of convection are not correlated with the intensity fluctuations (not shown).
PV (PVU, shaded) at 1532 m height for selected times and vertical velocity (1 m s
Previous studies on vacillation cycles have used PV as a metric to show the structural changes of the vortex during the weakening and strengthening phases.
Azimuthally averaged PV (PVU, shaded) as a function of radial distance and height for selected times. The height-dependent RMW is indicated by the grey line. Also shown are the 1 m s
Figure
During the weakening phases there is a trend for the PV structure to become less ring-like. At the end of each weakening phase the trend suddenly reverses and the vorticity structure becomes more ring-like. The change in the tendency of the vorticity structure is very sudden and coincides exactly with the start and end of each phase. However, as indicated by Fig.
Change in PV over the past hour due to advection only (shaded, PVU h
To test whether PV transport between the eye and eyewall is occurring, Fig. The PV tendency due to the physical processes has also been calculated, with the cloud rebalancing term (PV change due to cloud formation) dominating. Overall, the PV change due to physical processes is large and positive on the inner side of the eyewall and responsible for the maintenance of the PV ring structure to counterbalance the PV loss due to vertical upward transport.
In addition to the radial PV structure the PV also varies azimuthally with the intensity fluctuations. One way of describing the azimuthal PV symmetry is the method of
Figure
To attempt to explain the causes of the change in PV structure the barotropic conversion rate was computed as in
The barotropic conversion rate describes how kinetic energy is transferred between eddies and the mean flow.
Figure
In order to understand the role of these isolated regions of deep rotating convection in the intensity fluctuations, their strength and prevalence prior to and during the weakening phases will be examined, particularly in their appearance as a manifestation of cooperative barotropic and convective instability. The involvement of the isolated regions of deep rotating convection as a potential trigger for the weakening will also be investigated.
Perturbation vertical velocity (m s
During the strengthening phases, isolated regions of deep rotating convection are apparent as small-scale local regions of high vorticity and vertical velocity within the eyewall. These features resemble vortical hot towers (VHTs), formally defined in
Figure
During the weakening phases, isolated regions of deep rotating convection rarely form such that in the middle of a weakening phase it is unusual to see one of these structures. The
The spin-up of a TC can be examined in terms of the tangential wind budget that describes contributions to the mean tangential wind tendency from radial and vertical advection, which can be further split up into mean and eddy contributions. A form of the tangential wind budget based on
In order to understand the contribution of the isolated regions of deep rotating convection to the spin-up or spin-down of the TC, the eddy and mean contributions to the tangential wind budget were examined. Figure
Colour shading shows the
Just above the boundary layer the eddy term has a positive contribution to the tangential wind budget in both S1 and W1 (Fig.
As with Fig.
However, the radial location of the isolated regions of deep rotating convection seems to be important; the isolated region of deep rotating convection inside the RMW in Fig.
To understand how the convective structures change with the intensity fluctuations the diabatic heating profiles are investigated, in particular, how the heating profiles change from strengthening phases transitioning to weakening phases. Understanding the distribution of the diabatic heating and its vertical and radial gradients can allow links to be made with the barotropic structure, through the spatial gradient of diabatic heating term in the PV generation equation. The diabatic heating (Figs.
Diabatic heating (shading, K h
During both weakening and strengthening phases there are some similarities, notably two separate heating maxima, one in the inflow boundary layer at around 1 km and the other in the mid-troposphere associated with the latent heat release above the freezing level in the free vortex at around 7 km. The majority of the heating occurs around the RMW in the eyewall, although small amounts of heating also occur out to 150 km associated with outer rainbands.
One of the biggest differences between the weakening and strengthening phases is the radial extent of their respective azimuthally averaged heating distributions. All of the weakening phases have a heating distribution with a greater radial extent compared to all of the strengthening phases (not shown). This can also be seen in the observations in Fig.
Diabatic heating (K h
The effect of eddy diabatic heating was also investigated. These results are not shown since the azimuthally averaged eddy heating was small, typically an order of magnitude smaller than the mean heating terms, which is similar to the results of, for instance,
In terms of how the heating distribution changes just prior to a weakening phase, Fig.
Over the next few hours the secondary convective ring becomes more symmetrical, and the isolated regions of deep rotating convection continue to become less visible. Eventually by
It was found that weakening phases were associated with weaker heating outside of the RMW compared to strengthening phases associated with stronger narrower columns of diabatic heating just inside the RMW, which is consistent with a simple balanced dynamical interpretation
If the boundary layer plays a significant role in the cause of the intensity fluctuations then it may be necessary to attempt to understand the fluctuations in terms of the boundary layer spin-up mechanism as described by
The agradient wind is the deviation of the tangential wind from gradient wind balance
Figure
Throughout the storm's lifetime the tangential wind is supergradient near the eyewall within the boundary layer, with the highest agradient wind being around 670 m. The supergradient wind is advected vertically upwards; above the boundary layer the radial outflow removes more absolute angular momentum than is gained by the vertical advection, so the wind is near-gradient wind balance. Above the boundary layer, the storm can intensify in two ways described in
Just prior to the weakening phase the inflow in the boundary layer at a radius of 35 km decreases (Fig.
The reduction in the boundary layer inflow from the decrease in the PGF is not enough to spin down the boundary layer, and the frictionally induced inflow remains strong. Therefore, at the surface, the reduction in maximum total winds (black line in Fig.
During the weakening phase an increase in the agradient wind is seen within the boundary layer (Fig.
The start of a strengthening phase is characterised by a strong outflow jet and a slightly subgradient “overshoot” (red line in Fig.
A key feature that appears during the weakening phases is a thin layer of outflow above the boundary layer which has been noted to occur in order to return the unbalanced supergradient tangential flow to gradient wind balance above the boundary layer. Another contributing factor to this outflow layer is a mismatch in the mass flux expelled from the boundary layer and ventilated by the deep convection. The residual mass that cannot be evacuated through the main-system-scale tropospheric outflow channel must leave through the outflow jet at the top of the boundary layer. In order to better understand the changes in the strength of the outflow jet and its importance in causing weakening phases, the ventilation diagnostic as developed in
Figure
To understand how the boundary layer and outflow jet change and lead to a spin-down above the boundary layer, Fig.
Panels
The increase of the agradient wind at the start of the weakening phase leading to an intensification of the outflow jet can be seen by comparing Fig.
In summary, the intensity fluctuations in Hurricane Irma can be understood in terms of unbalanced boundary layer dynamics and the interplay between the boundary layer and the free vortex above. Firstly, the agradient wind in the boundary layer increases as a result of a decline of the PGF which is, itself, caused by an initial decrease in the azimuthally tangential wind above the boundary layer. The rapid increase in the supergradient wind within the boundary layer is associated in part with an intensification of the outflow jet just above the boundary layer which acts to spin down the primary circulation above the boundary layer by advecting in low angular momentum air from the eye, as well as expanding the RMW above the boundary layer. An increased supergradient wind also implies a stronger agradient force, promoting ascent out of the boundary layer at larger radii which can be seen explicitly by looking at Fig.
During the weakening phases the RMW expanded, the azimuthally averaged tangential wind speed (at all height levels in the lower and mid troposphere) decreased, and the MSLP stagnated or rose, whereas during the strengthening phases the opposite occurred.
The fluctuations observed in Hurricane Irma are proposed to be the result of changes in the barotropic structure (namely the proposed onset of barotropic instability during the strengthening phases) cooperating with convective instability to reduce the e-folding time of disturbances from barotropic instabilities similar to the arguments presented in The rise in barotropic instability is inferred by the growing wave-2 PV anomalies during the strengthening phase and the satisfaction of the Rayleigh–Kuo criterion, where a sign change is evident in the radial gradient of vorticity. However, in order to fully verify the existence and increase in barotropic instability a linear stability analysis would be a useful extension to this paper.
The start of a weakening phase during a rapid intensification period is associated with the presence of isolated regions of deep rotating convection built up over the course of the preceding strengthening phases that produce significant heating and are visible in an azimuthally averaged perspective as a secondary updraft outside of the RMW (Fig. Plotting the relative vorticity rather than the PV shows a similar tendency as in Figs. 6 or 7a, with the radial gradient of vorticity between the eyewall rapidly decreasing during the middle of the weakening phases but not to the extent that a “bridge” or monopole structure forms. Hence, the change in the PV distribution between the strengthening phase and the weakening phase is linked to thermodynamic structural changes as well as proposed barotropic stability changes.
This process seems to be similar to that described in the observational study of
The dynamical effect of the isolated regions of rotating deep convection may be an important element in the transition between the strengthening and weakening phases. During the strengthening phases these isolated regions of deep rotating convection are not harmful to the storm's intensification and can, through the eddy radial vorticity flux (term 3 on the right hand side of Eq.
The weakening that occurs in the tangential wind above the boundary layer is accompanied by a decrease in the PGF which is also transmitted through to the boundary layer. This decrease in the PGF (Fig.
A potentially similar kind of intensity fluctuation explored in
The fluctuations presented here in Hurricane Irma do show similarities to vacillation cycles, particularly with the simulation conducted in
The intensity fluctuations in Irma also have some similarities to a “partial eyewall replacement cycle” described in
The prior analysis has been carried out for one ensemble forecast. To demonstrate the robustness of the analysis, composites of selected key results will be presented across multiple ensemble members. Five out of 18 ensemble members (including ensemble member 15), initialised on 3 September 00:00 UTC, showed the intensity fluctuations previously discussed. A further six ensembles also showed similar but weaker fluctuations. An additional model simulation, initialised on 2 September 12:00 UTC, found seven out of 18 ensemble members with the same kind of fluctuations. The following composites are based on the five ensemble members initialised on the 3 September at 00:00 UTC that show the strongest fluctuations. The composites are over all of the weakening and strengthening phases in all of these five ensemble forecasts. These weakening and strengthening phases vary in length from 1 to 10 h, with 4–5 h being typical. There are a total of 45 weakening and strengthening phases averaged over.
One of the key aspects of the analysis is the transition during weakening phases from a ring-like PV distribution at the start of the weakening phase towards a more bridge like PV distribution towards the end of the weakening phase. Figure
Composite PV tendencies (PVU h
Figure
Absolute angular momentum budget composites showing
The composites demonstrate that similar processes are likely occurring in the other ensemble members. The fluctuations in intensity that occurred during rapid intensification are not just limited to a single ensemble member. This study focuses on a single case, Hurricane Irma (2017), so it is unclear how common this type of intensity fluctuation is in TCs. The ensemble forecasts showed no link between the likelihood of the intensity fluctuations and the environmental conditions so the causes of the fluctuations are likely stochastic in nature (in particular with respect to the radial location of isolated regions of deep rotating convection that develop). The fluctuations are shown to occur in around a third of the ensemble forecasts suggesting they may be a common feature in rapid intensification and motivating analysis of more cases.
The main aim of this study was to determine the cause of the observed intensity fluctuations in Hurricane Irma (2017) during rapid intensification and to identify the processes responsible. Understanding these fluctuations is important as they can affect both the intensity and size of the RMW in the short-term and therefore the destructive potential of the TC. Although the intensity fluctuations have been observed at the surface (see Fig.
Schematic outlining the proposed mechanism for the fluctuations modelled during the rapid intensification of Hurricane Irma during
A summary of the key findings and interpretations is as follows:
In Hurricane Irma during the second period of rapid intensification, the focus of this study, intensity fluctuations occurred, defined as short-term intensification and weakening periods at 10 m height. During the weakening phases, MSLP increased, 10 m total wind speed decreased or remained constant, and the 10 m RMW increased. In contrast, during strengthening phases, MSLP decreased, 10 m total wind speed increased, and the 10 m RMW decreased. Isolated regions of rotating deep convection form stochastically during the strengthening phases (Fig. The effect of the isolated regions of rotating deep convection is to initially spin up the azimuthally averaged tangential wind in the eyewall region, above the boundary layer, with eddy vorticity flux and eddy vertical advection (terms 3 and 4 on the right hand side of Eq. During the strengthening phases, especially at 1500 m height, the radial P distribution is an elongated ring (i.e. more eccentric, see Fig. At the start of the weakening phase (Fig. By the middle of the weakening phase (Fig. The decrease in the azimuthally averaged tangential wind, above the boundary layer, can be linked to the strengthening outflow jet through the radial vorticity flux from the eye (region highlighted by the yellow circle in Fig. During the entire weakening phase, the storm centre MSLP rises nearly concurrently with the weakening of the maximum 10 m total wind. This finding is the opposite of The tangential wind above the boundary layer eventually starts to increase again (Fig.
In conclusion, the findings from this analysis, as summarised in Fig.
The agradient wind is determined by taking the gradient wind balance, where the pressure force is balanced by the sum of the Coriolis and centrifugal forces
Physically the agradient wind represents the deviation of the primary circulation from gradient wind balance. A subgradient wind means the wind speed is lower than the gradient wind, while a supergradient wind is higher than the gradient wind. In the boundary layer, both subgradient and supergradient winds are often found. At the surface, friction reduces the tangential wind and causes it to be subgradient, but the frictionally induced inflow can also lead to tangential acceleration at higher levels and smaller radii which sometimes results in a supergradient layer.
Observational data used in this paper are made available online by the Hurricane Research Division and are available at
WT: conceptualisation, formal analysis, investigation, methodology, software, writing – original draft preparation. JS: conceptualisation, supervision, writing – review and editing. AR: conceptualisation, supervision, writing – review and editing. CJS: data curation, methodology, supervision, writing – review and editing.
At least one of the (co-)authors is a member of the editorial board of
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We would like to especially thank Roger Smith and Michael Montgomery for devoting large amounts of their time and support in helping to improve rigour and accuracy. We would also like to thank the anonymous reviewer for their helpful suggestions and comments. We thank the Hurricane Research Division for providing dropsonde and flight-level data as well as the images contained in Fig.
William Torgerson was funded by a PhD scholarship from the NERC SPHERES DTP (grant NE/L002574/1) and CASE support from the Met Office. This research was also partially funded by the Met Office Weather and Climate Science for Service Partnership (WCSSP), Southeast Asia, as part of the Newton Fund.
This paper was edited by Juerg Schmidli and reviewed by Roger Smith and one anonymous referee.