Appraisal of Data Assimilation Techniques for Dynamical Downscaling of the Structure and Intensity of Tropical Cyclones

Abstract

The dynamical downscaling technique is used for the understanding of physical mechanisms associated with the atmospheric phenomena. We have developed high‐resolution analysis (6 km) for three tropical cyclones (TCs), namely, Phailin (2013), Nilofar (2014), and Chapala (2015) originated over the North Indian Ocean using the dynamical downscaling approach. The study aimed at the identification of appropriate methodology for generating analysis so that it becomes useful for identifying the role of environmental and internal dynamics on intensification processes and structural changes of TCs. The simulations using Weather Research and Forecasting model and four‐dimensional variational (4DVAR), hybrid three‐dimensional ensemble‐variational (3DEnVAR), and a hybrid four‐dimensional ensemble‐variational (4DEnVAR) data assimilation (DA) techniques are compared. The impact of DA is quantified by comparing errors in position, minimum sea level pressure, and maximum wind speed with the best track data set of India Meteorological Department. The intensities of TCs simulated by three downscaling methods are validated in terms of changes in minimum sea level pressure, maximum surface winds, and boundary layer and middle tropospheric relative humidity. The skills scores, namely, equitable threat score, false alarm ratio, the probability of detection, and biases (BIAS), are calculated to identify the best suitable DA technique. It is found that the hybrid DA techniques improve the overall quality of analysis compared to those developed using only variational DA techniques. The simulation using the hybrid 4DEnVAR DA technique is found to be better for simulation of the track, intensity changes, and structural characteristics of TCs.

Evaluation of Multiplicative Weight of Covariance Matrix on Hybrid Data Assimilation Schemes

“…Various foreign authors agree that giving a higher weight to the ensemble BEC, because theoretically its better represents the error of the day, leads to more realistic solutions [10] [11]. A similar result was obtained by Zhu et al [12], who estimated that a much higher weight for the ensemble covariance in a hybrid covariance, helps the proposed assimilation system to produce a more reasonable analysis field.…”

Understanding uncertainties in contemporary and future extreme wave events for broad-scale impact and adaptation planning

“…Maintaining such networks to enable climate relevant over time will provide data with greatly improved spatial resolution to complement the network of fixed observation platforms. Improving representation of TC and ETC systems within global atmospheric reanalyses with more data assimilation ( 63 , 64 ) and development of new globally downscaled atmospheric reanalysis data ( 65 ) would also help to further constrain uncertainties. Expanding the limited number of dynamical global wave simulations forced by high-resolution downscaled climate models (0.25° or less) that are more capable of resolving TC-driven waves (relative to coarse CMIP-driven global wave simulations) ( 53 , 54 ) might also help to understand and reduce uncertainties in ETC and TC areas.…”

A Sampling‐Based Path Planning Algorithm for Improving Observations in Tropical Cyclones

“…The impact of the internal dynamics on TC intensification and structural changes revealed that hybrid data assimilation techniques improved the overall quality of prediction compared to individual data assimilation methods (Malakar et al., 2020). A flow‐dependent sequential assimilation‐based targeted observation using the ensemble Kalman filter could minimize the analysis variance (Wu et al., 2020).…”