• Comparison of the performance of HYBRID ETKF-3DVAR and 3DVAR data assimilation scheme on the forecast of tropical cyclones formed over the Bay of Bengal

• # Fulltext

https://www.ias.ac.in/article/fulltext/jess/129/0233

• # Keywords

Data assimilation; tropical cyclones; Bay of Bengal; WRF; HYBRID 3DVAR.

• # Abstract

This study compares the performance of hybrid ensemble transform Kalman filter – three dimensional variational data assimilation (HYBRID) system and three dimensional variational (3DVAR) data assimilation system in Weather Research and Forecasting Model (WRF) in simulating tropical cyclones (TC) formed over the Bay of Bengal. An Ensemble Transform Kalman Filter (ETKF) system updates the ensemble system that provides flow-evolving background error covariance for HYBRID data assimilation system. Results indicate that use of flow-evolving ensemble error covariance in 3DVAR system has systematically reduced the TC position and intensity errors in the analysis; however, adding more weights to the ensemble error covariance term in 3DVAR cost function has not made any significant impact. The 3DVAR analysis depicts a stronger TC vortex with a well pronounced warm core structure as compared to that in HYBRID analysis. The forecasts from HYBRID analysis outperform that from 3DVAR in reducing TC track forecast error. The relative improvement in TC landfall position is 43% and 49% for variously configured HYBRID experiments. The forecasts initiated from HYBRID analysis has higher skill in quantitative precipitation forecasts during TC landfall compared to 3DVAR, which may be attributed to improved track prediction in the HYBRID experiments.

$\bf{Highlights}$

$\bullet$ Compared the performance of HYBRID and 3DVAR data assimilation system for Tropical cyclone forecasts.

$\bullet$ HYBRID has systematically reduced the Tropical cyclone position and intensity errors in the analysis.

$\bullet$ The forecasts from HYBRID analysis outperform that from 3DVAR in reducing TC track forecast error.

$\bullet$ The forecasts initiated from HYBRID analysis has higher skill in quantitative precipitation forecasts during Tropical cyclone landfall compared to 3DVAR.

• # Author Affiliations

1. Department of Earth and Space Science, Indian Institute of Space Science and Technology, Trivandrum 695 547, India.
2. North Eastern Space Application Centre, Meghalaya 793 103, India.
3. CSIR Fourth Paradigm Institute, Bangalore 560 037, India.

• # Journal of Earth System Science

Volume 130, 2021
All articles
Continuous Article Publishing mode

• # Editorial Note on Continuous Article Publication

Posted on July 25, 2019