A Study on the Structural Characteristics of North Indian Ocean Tropical Cyclones during 2001-2020

Abstract

The tropical cyclones (or TCs) being one of the most extreme weather events, have the potential to cause catastrophic destruction to both human lives and properties. Its life span generally extends from a few days to weeks and the destruction caused by its combined effect of duration and size is incomparable to any other forms of atmospheric disturbances. Recent findings have suggested the rise in intensified TCs and their structural modulation in the global ocean basins like North Atlantic (NA) and western North Pacific (WNP). Over the North Indian Ocean (NIO) basin, several observational and modeling studies have been carried out during past three decades, which included the climatological aspects. However, the structural characteristics involving TC size, radial parameters, vertical structure, and associated environmental conditions have not been emphasized extensively. To address the said aspect, the present study emphasizes upon the estimation and distribution of TC size, radial parameters, and related meteorological factors, by using the scatterometer observations, global analyses, IMD (India Meteorological Department) and JTWC (Joint Typhoon Warning Centre) best track records, besides other relevant data sets. Also, the study adopted Weather Research and Forecasting (WRF) modeling system and the three dimensional variational data assimilation technique through WRF-DA to understand the impact of scatterometer winds on the structural characteristics. The study period is constrained to 2001-2020, based on the availability of the scatterometer data sets at the time of execution of the work. Accordingly, observational studies considered 71 TCs, whereas the modeling study considered 67 cases. This study analyzes the mean annual variations of TC size and radial parameters in the NIO basin, measured in terms of the radius of closed isobars (ROCI), Rvor (i.e. radius of 1×10-5 s-1 contour), radius of maximum winds (Rmax) and critical wind radii of 34 (R34), 50 (R50), and 64 (R64) knots (1 knot = 0.514444 ms-1) wind, computed following the vorticity based approach. The results presented in this work indicate that the error between TC centers determined by the vorticity-based approach and IMD data is in the range of 30– 105 km (mean value is 76.34 km) and with JTWC, the range is within 25–115 km (mean value 79.77 km) when all of the observations are considered. However, if the observations with the smallest time gap of < 30 min are considered, the corresponding mean error values become 30.12 km (±20.83 km) and 36.47 km (±20.21 km). The comparison of Rmax with JTWC best track records yields a correlation value of ~0.75 with RMSE of ~7.6 km. The radial parameter Rmax varies in the range of 15–90 km, R64 is within 15–60 km, R34 lies in the range of 80–300 km, R50 varies within 15–60 km, and Rvor is within 190–485 km for the NIO region that is comparable with the JTWC data. Larger size TCs are observed near the higher latitude of the NIO region, and the size of systems increases with an increase in latitude. The size of the pre monsoon season TCs is found to be larger in comparison to post-monsoon season systems over NIO. An empirical relationship is also established based on the linear regression method to determine the size of TCs using SCATSAT-1 data. The results presented in this work indicate that scatterometer wind data with the spatial resolution of ~6 to 25 km would be handy enough to study the structural characteristics of NIO TCs. A Statistical analysis is conducted to detect the increasing or decreasing trends of TC size, accumulated cyclone energy (ACE) and power dissipative index (PDI) with environmental factors in the context of annual and seasonal variability. The possible causes of environmental factors influencing TC size are investigated. Results suggest the annual mean values of PDI, ACE and ROCI increase and the mean value of Rmax decreases in NIO with statistical significance of 95%. The present study reveals substantial agreement between TC size parameter, ACE, PDI and environmental factors. Furthermore, the impact of scatterometer observed winds is studied using WRF/WRF-DA modeling system. WRF/WRF-DA considered a two nested domain configuration, with a 30 km horizontal resolution of the parent and 6 km nested domain. Two types of simulations were performed, one without assimilation, but considers SST data, and another with assimilation of scatterometer winds through three-dimensional variational technique and consideration of SST. The first one is named as CTRL and the second one is called DASST. Both the simulations considered same set of physics and dynamics and were initialized with NCEP-FNL and NOAA-SST data sets. For the 67 NIO TC cases considered in the study, a total of 134 simulations were performed. Out of the considered 67 cases, 43 are BOB TCs and 24 are from AS sub-basin. Each case is simulated starting from 00 UTC and the duration of the model integration is decided by referring to the IMD best track data. However, the nested domain is considered over the respective subbasins, keeping the horizontal resolution unchanged, but altering the initial and final grid points and number of grid points. For analyzing the genesis of TCs over BOB, whole subbasin is divided into three sectors, viz., north (15°N– 22.5°N; NBOB), middle (10°N-15°N; MBOB) and south (5°N- 10°N; SBOB) sector. Similarly, the AS sub-basin is divided into two sectors i.e., eastern (55°E - 65°E; EAS) and western (65°E - 77.5°E; WAS) sectors. The results helped to devise empirical formulations valid for a particular sub-basin or the whole NIO basin. The empirical relations are formulated to estimate Rmax values in case of NIO TCs in terms of R50 and TC size. Similarly, two other empirical relations are also formulated for estimating R34 and MSW values in terms of R50 and TC size over the NIO basin. The result is shown by considering all simulations and JTWC values. A reasonably higher correlation (> 0.5) demonstrates that the new method effectively reduced the estimation error of Rmax. Also, the simulated vertical structures of the TCs are also analyzed. For this purpose, the cross sectional illustrations of relevant parameters are taken into consideration.