RESEARCH
Embedded Files
Research interests
Research interests
My Ph.D. research deals with understanding the downstream amplification of Indian summer monsoon low-pressure systems (LPS) using the global reanalysis, and outputs from climate models of Coupled Modeled Intercomparison Project (CMIP5 and CMIP6). I am currently working on the dynamical linkage between West-North Pacific (WNP) tropical cyclones (TCs) with the LPS. My broader research interests also include the dynamics of synoptic-scale systems, atmospheric wave interactions, and the application of AI/ML in Atmospheric Science.
Indian summer monsoon LPS activity
Indian summer monsoon LPS activity
1) Downstream and in situ LPS genesis in CMIP5 models
1) Downstream and in situ LPS genesis in CMIP5 models
Keypoints of the research:
We proposed an automated algorithm to classify LPS based on their genesis into downstream and in situ mechanisms.
Downstream and in situ LPS genesis in the CMIP5 models are identified for the first time.
A weaker Potential Vorticity advection causes a weaker and incoherent propagation of LPS in CMIP5 models.
The in situ LPS genesis are dominant in all models inline with the observations.
Fig. 1 June to September climatology of track density (units: no. of LPS per grid per season) and LPS propagation vectors (units: m s−1) for (a) European center interim reanalysis (ERAI) and (b) multi-model ensemble of CMIP5; storm centered composites of 500 hPa–300 hPa average potential vorticity (contours; units: PVU) and potential vorticity advection (shading; units: PVU day−1) for (c) ERAI, and (d) multi-model ensemble mean of CMIP5. The potential vorticity contours start at 0.36 and has an interval of 0.12. The latitudes and longitudes in (c) and (d) are relative latitudes and longitudes with respect to the LPS center. The calculations are done for the period 1979–2005.
Fig. 2 Lead lag composite of relative vorticity at 850 hPa (units: 10−5 s−1) for (a) downstream, (b) in situ, and (c) uncertain low-pressure system (LPS) genesis days during 1979–2017 tracked ERAI reanalysis; (d–f) same as (a–c) but for the ensemble mean relative vorticity composites for the LPS tracked from 11 CMIP5 models during 1979–2005. These figures are from Srujan et al. (2021) . Srujan, K. S. S. S., S. Sandeep, and E. Suhas (2021) Downstream and in-situ genesis of monsoon low-pressure systems in climate models, Earth and Space Science, doi: 10.1029/2021EA001741
2) Dynamical linkage between Western North Pacific tropical cyclones and LPS
2) Dynamical linkage between Western North Pacific tropical cyclones and LPS
Key points of the research:
Causality has been established between Western North Pacific (WNP) tropical cyclones (TC) and LPS through transfer entropy.
Landfalling TCs near the South China Sea coast and adjacent region triggers the downstream LPS over BoB.
Clustering of TCs is performed based on the geographical tracks, location of genesis, angle of recurving, etc., and it is found that 83% of the downstream LPS are associated with the four clusters of WNP TCs only.
NEWS REPORTS:
Machine Learning (ML) application in Atmospheric Science
Machine Learning (ML) application in Atmospheric Science
Prediction of synoptic-scale Sea Level Pressure (SLP) over the Indian monsoon region using deep learning
Key points of the research:
In this research, it was shown that sea-level pressure (SLP) can be used as a proxy to predict the active-break cycles as well as the genesis of LPS using the deep learning model (ConvLSTM).
ConvLSTM model is able to reliably predict the daily sea level pressure anomalies over central India and the Bay of Bengal (BoB) at a lead time of 7 days.
A comparison of the ConvLSTM model predicted SLP with the forecasts from the numerical weather prediction models shows that the deep learning model has a better skill in capturing the synoptic-scale SLP fluctuations over central India and BoB
Fig. 1 Time series of observed daily precipitation anomalies, observed SLP anomalies, and predicted SLP anomalies over continental India for the years (a) 2014, (b) 2015, (c) 2016, and (d) 2017. The horizontal lines indicate ±1 standard deviation of the daily precipitation anomalies during June 1–30, September. The active and break days of the monsoon are shaded in blue and red, respectively. The SLP anomalies are predicted at a lead time of 7 days. This figure is from Sinha et al. (2021).
Sinha, A., M. Gupta, K. S. S. S. Srujan, H. Kodamana, and S. Sandeep (2021) Prediction of synoptic-scale sea level pressure over the Indian monsoon region using deep learning, IEEE Geoscience and Remote Sensing Letters, doi:10.1109/LGRS.2021.3100899
Page updated
Google Sites
Report abuse