Simulations of Cyclone Sidr in the Bay of Bengal with a High-Resolution Model: Sensitivity to Large-Scale Boundary Forcing

Meteorol Atmos Phys (2011) 114:123–137 DOI 10.1007/s00703-011-0161-9

ORIGINAL PAPER

Simulations of Cyclone Sidr in the Bay of Bengal with a high-resolution model: sensitivity to large-scale boundary forcing

Anil Kumar • James Done • Jimy Dudhia • Dev Niyogi

Received: 13 July 2009 / Accepted: 19 August 2011 / Published online: 4 September 2011 Ó Springer-Verlag 2011

Abstract The predictability of Cyclone Sidr in the Bay of the 500-hPa level. Comparison of the high resolution, Bengal was explored in terms of track and intensity using moving nested domain with a single coarser resolution the Advanced Research Hurricane Weather Research domain showed little difference in tracks, but resulted in Forecast (AHW) model. This constitutes the first applica- significantly different intensities. Experiments on the tion of the AHW over an area that lies outside the region of domain size with regard to the total precipitation simulated the North Atlantic for which this model was developed and by the model showed that precipitation patterns and 10-m tested. Several experiments were conducted to understand surface winds were also different. This was mainly due to the possible contributing factors that affected Sidr’s the mid-latitude westerly flow across the west side of the intensity and track simulation by varying the initial start model domain. The analysis also suggested that the total time and domain size. Results show that Sidr’s track was precipitation pattern and track was unchanged when the strongly controlled by the synoptic flow at the 500-hPa domain was extended toward the east, north, and south. level, seen especially due to the strong mid-latitude wes- Furthermore, this highlights our conclusion that Sidr was terly over north-central India. A 96-h forecast produced influenced from the west side of the domain. The dis- westerly winds over north-central India at the 500-hPa placement error was significantly reduced after the domain level that were notably weaker; this likely caused the size from the western model boundary was decreased. modeled cyclone track to drift from the observed actual Study results demonstrate the capability and need of a track. Reducing the model domain size reduced model high-resolution mesoscale modeling framework for simu- error in the synoptic-scale winds at 500 hPa and produced lating the complex interactions that contribute to the for- an improved cyclone track. Specifically, the cyclone track mation of tropical cyclones over the Bay of Bengal region. appeared to be sensitive to the upstream synoptic flow, and was, therefore, sensitive to the location of the western boundary of the domain. However, cyclone intensity 1 Introduction remained largely unaffected by this synoptic wind error at Accurate cyclone track and intensity predictions remain a challenging task for atmospheric scientists and the research Responsible editor: C. Simmer. community. A large number of cyclones form in the Bay of A. Kumar J. Done J. Dudhia Bengal (hereafter BOB) region and make landfall along the National Center for Atmospheric Research, Boulder, CO, USA coastal regions of India, Bangladesh, and Myanmar. These cyclones have been responsible for billions of dollars in A. Kumar D. Niyogi Purdue University, West Lafayette, IN, USA property damage, loss of agriculture crops, and thousands of human lives (e.g., Paul 2010). Between October and Present Address: December, cyclonically favorable, large-scale atmospheric & A. Kumar ( ) conditions are typical over BOB. Hydrological Science Branch, NASA/GSFC, Code-614.3, Greenbelt, MD 20771, USA This study concerns the simulation of a recent, notable e-mail: [email protected] BOB storm—Cyclone Sidr using the Advanced Research 123 124 A. Kumar et al.

Hurricane Weather Research Forecast (AHW) model. This and damage) to strike Bangladesh since 1991. JTWC would be the first test of the AHW model (Davis et al. issued a forecast on 9 November 2007 identifying a trop- 2008; Xiao et al. 2009) outside the Atlantic basin or the ical disturbance with weak, low-level circulation near the region for which it was developed and evaluated. We chose Nicobar Islands. Initially, a moderate upper-level wind Sidr mainly because of the (a) very strong wind and Saffir– shear with strong diffluence aloft aided in the developing Simpson equivalent category 5 intensity associated with convection zone. The vertical shear decreased greatly as this cyclone, (b) most of the operational models used for the circulation became better defined. As a result, a tropical forecasting purposes failed to capture track as well as cyclone formation alert was issued on 11 November, at a intensity, and (c) to help clarify the influence of strong time when the circulation was located a short distance upper-level mid-latitude westerlies over north India on the south of the Andaman Islands. JTWC warnings were based simulated Sidr cyclone under different domain dimensions. upon both Windsat microwave images that showed a low- Cyclone Sidr has been used as a test case by a number of level circulation center and upper-level analyses that modeling groups. Pattanayak and Mohanty (2008) and then showed enhanced convection due to a strong diffluent flow Bhaskar Rao and Srinavas (2010) reported on the perfor- over the disturbance. Around the same time, the India mance of MM5 and the Weather Research Forecast (WRF) Meteorological Department (IMD) designated the system modeling system on track and intensity changes. They as a depression and issued a warning stating that a showed that there is no significant improvement in the ‘‘depression has formed over the southeast Bay of Bengal model after a 36-h model forecast, because the model and adjoining Andaman Sea and lay centered at 1430 hours boundary and initial conditions provided by the coarser IST (India Standard Time) of 11 November 2007 near resolution NCEP forcing data dominated the results. Sim- 10.0°N and 92.0°E about 200 km south–southwest of Port ilarly, Badarinath et al. (2009) used the Sidr case with the Blair and the system is likely to intensify further and move MM5 model to assess aerosol loading. Kotal et al. (2008) in a west north mid-latitude westerly direction.’’ Figure 1a tested a statistical–dynamical approach to understand the shows the tracks that were issued every 6 h by the JTWC, errors in the Sidr track and found a northwest directional the US National Hurricane Center (NHC), and the Central bias. More recently, Akter and Tsuboki (2010) simulated Pacific Hurricane Center (CPHC). The JTWC upgraded the supercells in the Sidr rainbands with a cloud resolving Sidr to a tropical cyclone after Dvorak estimates indicated model to understand the synoptic latent heat and storm axis winds of 65 km h-1 (40 mph) on 11 November. Moreover, interactions. as the day progressed, the storm intensified into a deep In the following section, Sidr’s track and intensity depression as it moved slowly northwestward. The track is changes are discussed. This is followed by the AHW model shown in Fig. 1a with the global sea surface temperature description in Sect. 3. Model performance and track anal- (RTG_SST) analysis at 00/11 November 2007, developed ysis are presented in Sect. 4. The impact of domain size on through the National Centers for Environmental Prediction/ model track is presented in Sect. 5. Details of the cyclone Marine Modeling and Analysis Branch (NCEP/MMAB). structural features of the core are given in Sect. 6. Study The IMD observed track is plotted in Fig. 1b with the conclusions are summarized in Sect. 7. intensity and track discussions. Figure 1a and b shows slightly different tracks: at 0600 UTC 15 November, IMD estimated 132.5 mph surface winds, whereas JTWC shows 2 Sidr description 135 mph. There is no surface wind speed data available from IMD after 1800 UTC 15 November, which causes us Cyclone Sidr was the fourth named storm of the 2007 to rely solely on JTWC track records for that information. northern Indian Ocean cyclone season. Sidr formed in the The cyclone intensified to reach peak winds of 132.5 mph central BOB region and quickly strengthened to reach at 0600 UTC 15 November based on IMD observations 1-min sustained winds of 225.3 km h-1 (150 mph), and agrees with the JTWC estimates of 135 mph peak wind according to the Joint Typhoon Warning Centre (JTWC). speed for the same time. Sidr officially made landfall at This report qualified Sidr as a category 5 equivalent trop- 1600 UTC 15 November as per IMD track (IMD 2008). ical cyclone on the Saffir–Simpson Scale from 06 UTC 15 November 2007. The storm eventually made landfall in Bangladesh on 15 November 2007. According to the media 3 Model description reports, the storm caused large-scale evacuations of about 650,000 people and resulted in more than 2,400 fatalities. The Advanced Hurricane WRF (AHW) is a derivative of Most of the deaths were attributed to falling trees that the Advanced Research WRF model. The model is capable flattened many coastal structures. Cyclone Sidr was of resolving multiscale cyclone features from about 1 km described as the most severe storm (in terms of fatalities to synoptic-scale feedbacks. The technical details are 123 Simulations of Cyclone Sidr in the Bay of Bengal 125

Fig. 2 Model, nested domains (at 12-, 4-, and 1.33-km resolution and terrain height in m)

with the center of the cyclone. The model used the WSM3 microphysics scheme (Hong et al. 2004), while the Rapid Radiative Transfer Model (RRTM, Mlawer et al. 1997) and the Dudhia scheme (Dudhia 1989) were used for the longwave and shortwave radiation calculations, respectively. The thermal diffusion scheme was used to represent surface physics with the Yonsei University (YSU) planetary boundary layer scheme (Noh et al. 2003). The initial and boundary conditions for the large-scale atmospheric fields were derived from the 1 91 degree NCEP global final analysis (FNL) using the WPS (WRF Pre-processing Fig. 1 a JWTC estimated track and spatial pattern of sea surface System) software package. The model run started at 00Z 11 temperature (from 1/12-degree real-time Global SST analysis) at 0000 UTC 11 November 2007 in the Bay of Bengal, and b IMD November 2007 and ended at 00Z 17 November 2007. Sea recorded observed track (no data is available after 1506 UTC in IMD surface temperatures were derived from the high-resolution source) real-time global sea surface temperature (RTG_SST) at 1/12-degree resolution analyses from NCEP/MMAB. available at the WRF repository (http://www.mmm. ucar.edu/wrf/users/docs/arw_v3.pdf). For our simulations, 3.1 Surface flux parameterization the outermost domain was fixed (Fig. 2) with 12-km grid spacing (423 9 324 grid points), two nested movable Hurricane intensity is sensitive to the parameterization of domains at 4 km (201 9 201 grid points), and a 1.33-km momentum and enthalpy fluxes between the surface and grid spacing (240 9 240 grid points) that covered an area the atmosphere (Rosenthal 1971; Emanuel 1995). In the of 320 km 9 320 km and was configured with a two-way storm core, maximum wind speed depends on the square nesting option. The choice of inner domain grid spacing root of the ratio of the drag and enthalpy exchange coef- follows the findings of Chen et al. (2007) that for the WRF ficients, ðC =C Þ1=2 following Emanuel (1986). The sur- model, proper treatment of the inner core requires a grid k D face drag parameterization in the AHW model is based on spacing of less than 2 km. All domains had 35 vertical Donelan et al. (2004) which defines the relation between layers with a terrain that followed sigma coordinates with roughness length (Z ) and frictional velocity (u ) as, the model top at 0.5 hPa. The nest positions were updated 0 * 1=3 every 15 min of the simulation and the track was updated Z0 ¼ 10 expð9=u Þ;

123 126 A. Kumar et al.

where Z0 is bound by a limiting range between 0.125 9 10-6 and 2.85 9 10-3 m, respectively. Further- more, the Ck formulation was modiﬁed with the so-called ramped Ck approach by introducing a ramping effect in the enthalpy roughness length as described in Dudhia et al.

(2008). This ramped Ck up with wind speeds of hurricane strength.

3.2 Coupling with a 1D ocean model

Ocean temperature feedback was applied to every grid point in the AHW model through a 1D ocean model based on Pollard et al. (1973). The ocean model was initialized for this case with a 30-m ocean mixed-layer depth (MLD). Rao et al. (1989) studied the mean monthly MLD in the Arabian and BOB regions and found it to be between 30 and 40 m in November. The NCEP Global Ocean Data Assimilation System (GODAS) showed a 25- to 35-m MLD for the same month and was considered appropriate. Fig. 3 Simulated 1.33-km resolution-based track and intensity from The model does not consider lateral heat transfer different model initialization time and observed track (white line) between individual ocean columns, so heat only propagated vertically. This model accounted for the Coriolis effect, but shown in Fig. 4. The track started to diverge from the actual there was no advection or pressure gradient. A MLD of track at 0000 UTC 13 November toward the northwest and 30 m produced a maximum cooling of about 3.1 K when continued simulating the incorrect track with later prediction considering a deep-layer lapse rate of 0.05 km-1 (Davis times. The model also simulated a slower moving cyclone by et al. 2008). Frictional velocity estimations were through up to 3° latitude at 0000 UTC 15 November when compared surface layer physics and net radiation. Surface ﬂuxes to the observed location (17.8°N, 89.2°E). accounted for thermal forcing as secondary forcing only with ocean thermal mixing being the primary forcing. The 4.2 Model track analysis initialized at 0000 UTC 12 atmospheric model called the ocean 1D column model at November 2007 every time step and also updated the SSTs. In an attempt to improve the predicted track and lag time, the model was initiated at 0000 UTC 12 November. The 4 Results modeled track, shown in Fig. 3, again diverges from the actual track toward the northwest direction with only a 4.1 Model track analysis initialized at 0000 UTC 11 small improvement on the simulation with the earlier November 2007 model initialization time. The predicted intensity reached 127.3 knots with a center pressure as low as 933.11 hPa at Figure 3 shows the model-predicted track from the simula- 16.21°N and 86.92°E for 0000 UTC 15 November corre- tion initialized at 0000 UTC 11 November. The model- sponding to a category 4 cyclone, but the displacement simulated results presented in this section are from a moving error at this time was 1.17° (128 km) from the actual nest at 1.33-km domain resolution. The model track deﬂec- cyclone position. With the later model initialization times, ted to the left of the observed track, and resulted in landfall on the model improved the track by 0.5° toward the east and the Orissa coast, which was far from the actual landfall also improved the position and timing of the cyclone. Still location. However, the model was able to capture Sidr’s the predicted intensity was slightly weaker (by 14 knots) in intensity reasonably well. This model simulation at comparison to the model simulation initialized at 0000 UTC 15 November was indicative of a category 5 0000 UTC 12 November 2007. The model overpredicted cyclone with maximum sustained winds of 141 knots. The the maximum sustained winds (by 12 knots) in comparison model-estimated maximum sustained winds refer to 10-m with observed data. In summary, changing model initiali- winds and minimum surface pressure of 931.6 hPa posi- zation time not only made a small difference to track and tioned at 14.80°N and 86.43°E. The model-simulated tem- timing, but also had an impact on intensity (Fig. 4). To poral evolution of intensity and minimum surface pressure is gain further insight into the impact of model initialization

123 Simulations of Cyclone Sidr in the Bay of Bengal 127

Fig. 5 Four domains (at 12-km resolution) used in the domain size sensitivity study

conditions. This experiment is discussed in the following section.

4.4 Model track analysis initialized at 0000 UTC 13 Fig. 4 Time series for central minimum sea level pressure (hPa) and November 2007 maximum velocity (knots) of cyclone for simulations (at 1.33-km resolution) beginning at different initialized times Figure 3 shows good track yet poor intensity for a simulation initialized at 0000 UTC 13 November. For the large model domain used in the study region, it is well known time on cyclone track and intensity, we conducted further that the large-scale processes in the model diverge from tests described in the next section. those in the boundary conditions (Denis et al. 2003). Therefore, it is plausible that for simulations with an earlier 4.3 Model track analysis initialized at 1200 UTC 12 initialization time the model has time to develop large- November 2007 scale errors that result in larger cyclone track errors. Ini- tializing the model closer in time to landfall limits the error Figure 3 shows that the predicted track for a simulation growth at large scales, which may be the reason for the initialized at 1200 UTC 12 November still diverged from improved model cyclone track. However, the simulation the actual track toward the northwest. The model simula- with improved track also predicted poor intensity. tion at 0000 UTC 15 November showed maximum sus- Thus, the results show a large variation in minimum sea tainable winds of about 101 knots, surface pressure at level pressure, intensity, and track. The results also suggest 958.9 hPa, with its location at 14.59°N and 87.81°E for the that the cyclone track is controlled by large-scale features ﬁrst 24 h. The model follows the observed track, but such as synoptic winds while intensity strongly depends on thereafter diverges toward the northwest. Model tracks both local and large-scale conditions. were approximately the same as those seen in the simulation initialized at 0000 UTC 12 November, yet the maximum sustained winds dropped from 127 to 101 knots, 5 Impact of domain size on cyclone track while surface pressure increased from 933.11 to 958.9 hPa (Fig. 4). One possibility for the reduction in maximum To further assess the large-scale/local-scale interactions, winds may be a cooled sea surface. It was thought that we ran simulations initialized at 0000 UTC 11 November initializing the model later in time may improve track and out to 144 h for four different domain sizes, each at 12-km intensity due to more realistic lateral and boundary grid spacing. The domain sizes are shown in Fig. 5.This

123 128 A. Kumar et al.

Fig. 6 Predicted track using different domain sizes (at 12-km resolution) with model start time at 0000 UTC 11 November. DS1 is the largest domain, DS2 is the second largest, DS3 is the third largest, and DS4 is the smallest domain size as shown in Fig. 5

Fig. 8 Wind speed and vector (m s-1) at the 500-hPa level at 0000 UTC 15 November in domain 1 (at 12-km resolution), a model analysis, and b difference ﬁeld NCEP minus model

section discusses the model results corresponding to the 12-km resolution domain. The largest domain, domain 1, had 424 9 325 grid points (longitude: 65–115°E, latitude: 1°S–35°N). The second largest domain, domain 2, had 364 9 285 grid points (longitude: 70–110°E, latitude: 0–30°N). The third largest domain, domain 3, had 264 9 215 grid points (longitude: 75–105°E, latitude: 5–28°N), and, the smallest domain (domain 4), shown in Fig. 5, had 164 9 185 grid points (longitude: 80–98°E, latitude: 6–25°N). All grids were measured in west–east and north–south directions. Figure 6 shows the model track produced by the different domain-sized simulations. As the domain size decreased, the track improved. The simulation Fig. 7 NCEP FNL analysis (interpolated to 12-km resolution from 1° 9 1°) wind speed and vector (m s-1) at the 500-hPa level at on the smallest domain, domain 4, simulated a reasonable 0000 UTC 15 November 2007 track. Although there are only small differences in model 123 Simulations of Cyclone Sidr in the Bay of Bengal 129

Fig. 9 a Model-simulated (at 12-km resolution) wind speed and wind vector at the 500-hPa level in domain 2, b difference NCEP FNL analysis (shown in Fig. 6) minus model winds in domain 2 Fig. 10 Same as Fig. 9a, b except in domain 3 track between the third and fourth domain sizes, the track corresponding model-simulated wind vector and wind using the smallest domain (fourth domain size) is in good speed for domain 1 while the difference field is shown in agreement with the actual track during and after landfall. Fig. 8b. The analysis time of 0000 UTC 15 November was The track is slightly better for domain 3 than for domain 4 chosen because at this time there were significant differ- (i.e., the smallest domain track) until landfall. After land- ences among the tracks from various domain sizes when fall, the cyclone’s low pressure center moved into the model forecast time is 96 h. Figure 7 shows strong north- BOB. Since this opposes the observed track, domain 3 westerly (NW) winds over central India in the NCEP data cannot be considered a good track. The variability of while the model-simulated winds at the 96-h forecast time cyclone intensity with model domain size is discussed in show a generally weak flow. The field shows wind speed the next section. differences of 8–10 m s-1. Hence, for the largest domain, We used NCEP/NCAR FNL data to verify large-scale the model errors are largest for the 500-hPa NW flow over flow in the model. First we focus on the mid-latitude central India during the 96-h forecast. One possible reason westerly flow over northern India of winds at the 500-hPa is that the distances to the boundary conditions, which are level in order to find differences between the various approximately 1,200 km from central India in the west as domain-sized simulations with regard to their impact on well as in the east direction, allow freedom within the cyclone track. Figure 7 shows a FNL analysis wind vector interior of the domain for the model to diverge at a large and speed at 0000 UTC 15 November. Figure 8a shows the scale from the driving analysis. A similar analysis is

123 130 A. Kumar et al.

Fig. 12 a Model-simulated (at 12-km resolution) wind speed profile over Raipur station at 0000 UTC 15 November, and b temporal wind speed comparison at 500-hPa level over Raipur station Fig. 11 Same as Fig. 10a, b except in domain 4 carried out for the second largest domain, domain 2. Fig- ure 9a shows the model field while the difference field is updated every 6 h, provided stronger control on the large presented in Fig. 9b. The modeled 96-h forecast field scales in the model, which helped to improve the 500-hPa shows weak synoptic winds that are shifted slightly toward level westerlies as well as the cyclone track. Figure 11a the north. The difference field shows 12–15 m s-1 devia- shows model-simulated winds on the smallest domain, tions in wind speed over east-central India. Figure 10a domain 4, which has similar patterns and magnitudes at shows model-simulated winds for domain 3 with large scales as the reanalysis has. Differences from the improvements seen in a confined region of strong NW flow driving analysis reach 5 m s-1 as shown in Fig. 11b. in this domain. The difference field, Fig. 10b, shows values Domain 4 is small enough to capture the synoptic pattern of 6–10 m s-1 wind speed difference as well as differences over the BOB region and parts of Central India, Bangla- in wind vector fields. The model boundary forcing on the desh and Myanmar, and also sufficiently resembles the west side of the domain is located at 75°E longitude which observed cyclone track. We caution that our results should is approximately 500 km away from the strong mid-lati- be considered true only for such cases where strong syn- tude westerlies that is confined over central India. The optic flow influences a cyclone and may not be true for all closer proximity of the lateral boundary conditions, cases over BOB. 123 Simulations of Cyclone Sidr in the Bay of Bengal 131

Fig. 13 Model-simulated (at 12-km resolution) total precipitation (mm) observed between 0000 UTC 11 November and 0000 UTC 16 November and simulated wind barb (m s-1) at 0000 UTC 16 November is plotted for a domain 1 b domain 2, c domain 3, and d domain 4. For domain information see Fig. 5

The vertical wind speed profiles at 0000 UTC 15 flow on the smaller domains is more strongly constrained, November over Raipur station (21°N, 82°E) is shown in which results in an improved track prediction. Fig. 12a. We chose Raipur station mainly because it is The improvement on the smallest domain is explored situated under the region where we see significant changes further by extending domain 4 toward the east, north, and in wind speed at the 500-hPa level in the four different south directions by 5°. The model was run for this domain domain-sized experiments. We anticipate that the winds with nested inner domains at 4 and 1.33 km. One of the over broader central India including the Raipur region may objectives of this experiment was to determine the impor- be affecting the model’s lateral boundary from the west tance of the location of the western domain lateral side which can cause a track deflection despite the wind boundary condition that controls the simulated track and direction remaining the same in all four domain-size sim- compare those findings to the importance of domain size. ulations. The observed wind speeds were obtained from the The simulation for the extended domain did not show any Wyoming atmospheric sounding data archive and were significant difference in track (not shown). Therefore, this compared with modeled wind speed profiles of the differ- further highlights that model track is controlled by the west ent domain-size simulations. As expected, the wind profiles domain boundary flow conditions rather than domain size using the smaller domain sizes (domains 3 and 4) are or other lateral boundaries. The peak intensity of closest to the observed wind speed profile, especially 120 knots, however, was not maintained for long in the around the 500-hPa level. The temporal evolution of wind simulation. speed at 500 hPa in the observation and model-simulated Simulated total precipitation patterns from the different winds is shown in Fig. 12b, for the location (21°N, 82°E) domain-size experiments can be helpful in understanding where NW winds are strong. Winds in the smallest domain the impact of domain size on the model’s output. Figure 13 more closely follow the observed winds at the 500-hPa shows the total precipitation that occurred between level. With this analysis, we concluded that the interior 0000 UTC 11 November and 0000 UTC 16 November

123 132 A. Kumar et al.

Table 1 Displacement error, this error is calculated using center of observed cyclone at 24 h interval Synoptic time (UTC) Displacement error (km) YYYYMMDDHHMM DOM-1 DOM-2 DOM-3 DOM-4

200711160000 725 592 488 210 200711150000 260 225 188 190 200711140000 165 156 120 125 200711130000 112 72 55 58 200711120000 115 95 40 60 200711110000 70 65 62 62

Table 2 Date, location, and wind speed (mph) based on observed JWTC (Fig. 1a) Synoptic Time (UTC) Latitude Longitude Wind speed (mph) (YYYYMMDDHHMM)

200711160000 25.0 91.9 105 200711151200 22.8 90.3 130 200711151200 20.9 89.5 130 200711150600 19.3 89.3 135 200711150000 17.8 89.2 130 200711141800 16.6 89.3 130 200711141200 15.7 89.3 130 200711140600 15.0 89.4 120 200711140000 14.3 89.6 115 200711131800 13.7 89.5 115 200711131200 13.0 89.6 115 200711130600 12.5 89.8 115 200711130000 12.1 89.8 115 200711121800 11.6 90.0 105 200711121200 11.0 90.3 75 200711120600 10.8 90.4 55 200711120000 10.4 90.8 50 200711111800 10.4 91.4 45 200711111200 10.2 91.9 35 200711110600 10.0 92.3 35 Fig. 14 Model-simulated (at 12-km resolution) total precipitation (mm) observed between 0000 UTC 11 November and 0000 UTC 16 November and simulated wind barb (m s-1) at 0000 UTC 16 November is plotted for a experiment with domain 4, and b model estimates. We also investigated precipitation patterns from experiment with extended domain toward east, north and south with reference to domain 4 an extended domain experiment which expanded domain 4’s area toward the east, south, and north (Fig. 5). Fig- 2007 (120-h rain) along with the 10-m surface wind barb. ure 14a shows accumulated precipitation from 11 to 16 The heaviest precipitation was around the eyewall of the November 2007 in domain 4 while Fig. 14b shows the storm and followed the track. As seen in Fig. 13, the pre- accumulated precipitation for the same period in the cipitation patterns are signiﬁcantly different in each of the extended domain. Overall precipitation patterns were the four domain-size simulations and heavy precipitation fol- same and followed the heavy rain along the cyclone track. lows the model track. Figure 13d shows that projected This then conﬁrms that the east side of the domain rainfall was close to the high-resolution global precipita- boundary was not controlling the Sidr track. The dis- tion map (not shown) based off the satellite TRMM-PR placement error of the simulated cyclone eye location was

123 Simulations of Cyclone Sidr in the Bay of Bengal 133

Fig. 15 Model and observed MSLP and TC intensity comparison. Observed MSLP and TC intensity is denoted as ﬁlled triangle and circle, where as model estimate shows in open triangle and circle. Model results are from the simulation conducted using domain size 4 at 1.33-km resolution also calculated from domain-size experiments and is shown in Table 1. Incidentally, the cyclone eye displacement error was less in domain 3 and domain 4 than in domains 1 and 2. Domain 4’s displacement error was lowest during the 0000 UTC 16 November 2007 simulation. On the whole, a 210-km displacement error was found among all three domain sizes.

6 Sidr eyewall structure and intensity investigation

Fig. 16 a Single domain (at 12-km resolution) and three nested To examine Sidr more closely in terms of intensity and domains (at 1.33-km resolution) simulated tracks versus observed, minimum sea level pressure (MSLP), we compared the and b time series of minimum central pressure from single and nested intensity and central pressure obtained from the JWTC site model results and compared with observation with the domain 4 simulation (Table 2). Figure 15 shows close agreement for MSLP and maximum winds. After 00 UTC 14 November (a 72-h model forecast), the model drift/IR/WV winds, IR-proxy winds and Scatterometer predicted higher MSLP (940 hPa) than the satellite-derived winds, QuikSCAT, and Advanced-Scatterometer (A- value of 920 hPa pressure. Also, the model-estimated SCAT). A variational approach described in Knaff and maximum winds at 120 knots, whereas the satellite-derived DeMaria (2006) in conjunction with these five data sources estimates were closer to 140 knots. This confirms that the were used to create a mid-level (near 700 hPa) wind. Two model can predict intensity and MSLP reasonably well dimensional (2D) flight winds are estimated from IR while maintaining a good track throughout the 144-h imagery (Mueller et al. 2006). These 2D winds were forecast duration. obtained following AMSU derived wind fields and are used We also conducted an experiment with a single domain in solving the non-linear balance equations as described in and compared it to a nested domain simulation for domain Bessho et al. (2006). Figure 17a and b shows a comparison 4. The simulated tracks and minimum central pressures between satellite-derived winds and model-predicted winds presented in Fig. 16a and b showed that the single domain at 4-km grid resolution at 0000 UTC 15 November 2007 simulation displayed an improved track but had poor (96-h model forecast). The model-predicted wind direction intensity. This indicates that the simulated track was less at the 700-hPa level (Fig. 17c) was within reasonable dependent on the domain size and the simulated intensity agreement of the satellite-derived wind direction. It was of the cyclone is more dependent on model grid resolution. also noted that during the 96-h model forecast, at 0000 Model winds at the 700-hPa level were compared with UTC 15 November, the displacement error was 190 km satellite-based wind analysis (at 700 hPa). Winds estimates and the model-simulated wind near the periphery was include satellite data product reference datasets from the 50 knots in the domain 4 simulation conducted with a 1.33- Advanced Microwave Sounding Unit (AMSU), Cloud- km inner nested domain. This is close to the satellite- 123 134 A. Kumar et al.

Fig. 17 Wind speed at 0000 UTC 15 November for a satellite-estimated winds, b model winds in a movable nest domain at 4-km resolution, and c model wind vectors

derived winds of 60 knots (Fig. 18a, b). Due to a lack of be in good agreement in terms of band and overall cyclonic good quality data in and around the inner cyclone core, structure. further verification is limited. Satellite-derived winds suggested the eye was circular and symmetric in nature, yet the model predicted the storm 7 Conclusion eye was neither symmetric nor circular. To visualize the model-predicted storm eye, potential vorticity parameters In this study, we applied the AHW modeling system for an were used at different model forecast times. At 0000 UTC intense tropical cyclone in the Bay of Bengal region. The 14 November, the storm’s eye shape was both circular and study investigated the impact lateral and boundary forcing symmetric (Fig. 19). At later times, for instance, at 14 of four different domain sizes has on cyclone track and November 1200 UTC and 15 November 0000 UTC, the intensity and found that the reduction in domain size both eye took on a more triangular shape. Furthermore, at minimized the substantial model error growth in synoptic 1200 UTC 15 November, the eye was oval shaped. A winds and improved the cyclone track and storm intensity similar triangular-shaped eye was documented in a mod- during a complete 144-h model forecast. Our analysis eling study of Hurricane Katrina (Corbosiero et al. 2008). showed that the model simulated cyclone Sidr’s track was To view cyclone structure and associated bands, side-by- significantly influenced by a large-scale 500-hPa level mid- side comparisons of satellite IR imagery and model cloud latitude westerlies relative to flow on south and east of the top temperature (not shown) were made and were found to domain. For the large domain, the model underestimated

123 Simulations of Cyclone Sidr in the Bay of Bengal 135

large domain size showed that later initialization times not only improved the model-predicted track, but also produced poor cyclone intensity. To understand the relative importance of the location of the western boundary versus the domain size when predicting cyclone track, we extended the domain toward the east, north, and south directions and kept the western boundary the same. Results indicated that the simulated track and intensity are reasonable and hence, confirmed that the western boundary played a significant role in controlling the track. Our analysis also highlighted the impact of model domain resolution on track and intensity. Furthermore, it was found that with coarser resolution, the model predicted a good track but failed in terms of intensity. The domain size also affected the total simulated precipitation patterns, but the precipitation amount was not much different in different domain-size simulated experiments. Extending the domain toward the east, north, and south did not affect the simulated precipitation patterns, which implied that was only influenced by the westward large- scale boundary forcing. The displacement error in Sidr’s storm eye was significantly affected by changing the domain size used in modeling experiments, which implied that the displacement error decreased after reducing the domain size from west to east. Interestingly, the difference in displacement error between second smallest domain (domain 3) and smallest domain (domain 4) is small and many times domain 3 track is better by few kilometers but after making landfall, domain 3 simulated track is moved back in the ocean and get off the track completely on last day of simulation period. Hence, we made conclusion that smallest domain simulated track is better and follow actual track even after making landfall. For the smallest domain, where the model predicted both track and intensity in good agreement with observation, the model-predicted eyewall and structure was captured well. However, the triangular shape of the storm eye was not consistent with the more circular eye inferred from satellite-derived Fig. 18 Wind speed and direction along Sidr eye at 0000 UTC 15 November for a satellite-estimated winds, and b model-simulated winds and imagery. The model’s predicted storm loca- winds in a movable nest domain at 1.33-km resolution tion was generally within 150–200 km of the actual storm location. It is likely that the impact of domain size and boundary flow significantly affected the cyclone’s -1 wind speeds by 8–10 m s at the 500-hPa level over the motion at times when there was strong synoptic flow in central part of India resulting in a poor cyclone track this region, as seen here in the case of cyclone Sidr. projection. The underpredicted large-scale flow could be However, this may be less important for cyclones that corrected by reducing the domain size. This highlights the occur in weaker synoptic flows. importance of the mid-tropospheric flow for the tropical Our experimentation with domain size and analysis cyclone simulation. The reasons for the larger domains boundary conditions highlighted the importance of the failing to capture the feature accurately will need to be location of the western ‘inflow’ boundary. In forecast addressed. models, improvements may be possible in simulating the The main findings from this study are as follows. BOB cyclones by reducing the errors in the 500-hPa wind, Experimentation with model initialization time using the and the role of sounder data assimilation, better initial 123 136 A. Kumar et al.

Fig. 19 Model storm eye shapes at different times seen in the ﬁeld of potential vorticity at the 700-hPa level (1.33-km resolution). (1 PVU = 10-6 m2 s-1 Kkg-1) conditions, and improved model physics needs to be Bessho K, DeMaria M, Knaff JA (2006) Tropical cyclone wind investigated. retrievals from the Advanced Microwave Sounder Unit (AMSU): application to surface wind analysis. J App Meteorol 45:399–415 Acknowledgments The authors would like to thank Qingnong Bhaskar Rao DV, Srinivas D (2010) Real-time prediction of SIDR Xiao from MMM Division at National Center for Atmospheric Cyclone over Bay of Bengal using high-resolution mesoscale Research (NCAR) for the internal review on an initial draft. We models. Indian Ocean Trop Cyclones Clim Chang 3:159–167. also thank NCAR supercomputing resources for providing com- doi:10.1007/978-90-481-3109-9_20 puting GAUS. NCAR is sponsored by the National Science Chen SS, Price JF, Zhao W, Donelan MA, Walsh EJ (2007) The Foundation. The study also beneﬁted from the NSF CAREER grant CBLAST-Hurricane Program and the next-generation fully (ATM-0847472). coupled atmosphere–wave–ocean models for hurricane research and prediction. Bull Am Meteorol Soc 88:311–317 Corbosiero KL, Molinari J, Vollaro D, Wang W, Done JM (2008) The distribution of helicity and intense convection in tropical References cyclones. Paper P2F.13, 28th AMS Conference on hurricane and tropical meteorology, Orlando, 28 April–2 May 2008 Akter N, Tsuboki K (2010) Characteristics of supercells in the Davis C, Wang W, Chen SS, Chen Y, Corbosiero K, DeMaria M, rainband of numerically simulated Cyclone Sidr. SOLA Dudhia J, Holland G, Klemp J, Michalakes J, Reeves H, Rotunno 6A:25–28 R, Snyder C, Xiao Q (2008) Prediction of landfalling hurricanes Badarinath KVS, Kharol SK, Sharma A, Ramaswamy V, Kaskaoutis with the Advanced Hurricane WRF model. Mon Weather Rev DG, Kambezidis HD (2009) Investigations of an intense aerosol 136:1990–2005 loading during 2007 cyclone SIDR—a study using satellite data Denis B, Laprise R, Caya D (2003) Sensitivity of a regional climate and ground measurements over Indian region. Atmos Environ model to the resolution of the lateral boundary conditions. Clim 43:3708–3716 Dyn 20:107–126

123 Simulations of Cyclone Sidr in the Bay of Bengal 137

Donelan MA, Haus BK, Reul N, Plant WJ, Stiassnie M, Graber HC, Mlawer EJ, Taubman SJ, Brown PD, Iacono MJ, Clough SA (1997) Brown OB, Saltzman ES (2004) On the limiting aerodynamic Radiative transfer for inhomogeneous atmosphere: RRTM, a roughness of the ocean in very strong winds. Geophys Res Letts. validated correlated-k model for the long-wave. J Geophys Res doi:10.1029/2004GL019460 102:16663–16682 Dudhia J (1989) Numerical study of convection observed during the Mueller KJ, DeMaria M, Knaff JA, Vonder Haar TH (2006) Objective winter monsoon experiment using a mesoscale two-dimensional estimation of tropical cyclone wind structure from infrared model. J Atmos Sci 46:3077–3107 satellite data. Weather Forecast 21:990–1005 Dudhia J, Done J, Wang W, Chen Y, Xiao Q, Davis C, Holland G, Noh Y, Cheon WG, Hong SY, Raasch S (2003) Improvement of the Rotunno R, Torn R (2008) Prediction of Atlantic tropical k-profile model for the planetary boundary layer based on large cyclones with the Advanced Hurricane WRF (AHW) model. eddy simulation data. Bound Layer Meteorol 107:401–427 28th AMS Conference on hurricanes and tropical meteorology, Pattanayak S, Mohanty UC (2008) A comparative study on perfor- Paper 18A2, Orlando, 28 April–2 May 2008. http://ams.confex. mance of MM5 and WRF models in simulation of tropical com/ams/28Hurricanes/techprogram/paper_138004.htm cyclones over Indian seas. Curr Sci 95:7 Emanuel KA (1986) An air–sea interaction theory for tropical Paul B (2010) Human injuries caused by Bangladesh’s Cyclone Sidr: cyclones. Part I: steady-state maintenance. J Atmos Sci an empirical study. Nat Hazards 54:483–495 43:585–604 Pollard RT, Rhines PB, Thompson RORY (1973) The deepening of Emanuel KA (1995) Sensitivity of tropical cyclones to surface the wind mixed layer. Geophys Fluid Dyn 4:381–404 exchange coefficients and a revised steady-state model incorpo- Rao R, Molinari RL, Festa JF (1989) Evolution of the climatological rating eye dynamics. J Atmos Sci 52:3969–3976 near-surface thermal structure of the tropical Indian Ocean. Hong SY, Dudhia J, Chen SH (2004) A revised approach to ice J Geophys Res 94:1081–10815 microphysical processes for the bulk parameterization of clouds Rosenthal SL (1971) The response of a tropical cyclone model to and precipitation. Mon Weather Rev 132:103–120 variations in boundary layer parameters, initial conditions, lateral IMD (2008) Track of storm and depressions over the Indian Seas boundary conditions and domain size. Mon Weather Rev during 1891–2007, Cyclone e-Atlas of IMD. IMD, New Delhi 99:767–777 Knaff JA, DeMaria M (2006) A multi-platform satellite tropical Xiao Q, Zhang X, Davis C, Tuttle J, Holland G et al (2009) cyclone wind analysis system. http://rammb.cira.colostate. Experiments of hurricane initialization with airborne Doppler edu/products/tc_realtime/images/MPSW.pdf radar data for the Advanced-research Hurricane WRF (AHW) Kotal SD, Roy Bhowmik SK, Kundu PK (2008) Application of model. Mon Weather Rev. doi:10.1175/2009MWR2828.1 statistical–dynamical scheme for real time forecasting of the Bay of Bengal very severe cyclonic storm ‘‘Sidr’’ of November 2007. Geofizika 25:139–158

123