Modeling framework for thunderstorm/ lightning predictionThunderstorm/lightning Modeling Team, Monsoon Mission, IITM, Pune, India

Background

Proposed roadmap of

lightning/thunderstorm now-casting

Results: Case studies

Acknowledgements: Indian Institute of Tropical Meteorology, Pune is supported by the Ministry

of Earth Sciences, Govt. of India, New Delhi. Team sincerely thanks Director, IITM, Program

manager, Aaditya HPCS, WG-TP, MoES, IMD, NCMRWF.

RAC 2019

RAC meeting 2019: 05 - 06 February 2019, IITM, PUNE, INDIA

Thunderstorms (TS) are the source of lightning discharge, which is

the major cause of natural calamity (i.e. damage of public properties

and loss of life) across the globe. In India, the loss of human life

due to lightning strike is particularly high over different parts of the

sub-continent due to large occurrence of TS in pre-monsoon season

(March-May). Unfortunately, besides some purely empirical methods,

there is hardly any systematic mechanism involving dynamical model

and suits of observational inputs which provides a reliable forecast to

issue a warning prior to the occurrence of lightning.

A system of modeling framework for thunderstorm/lightning

prediction based on state-of the art dynamical model (e.g., WRF) as

well as hybrid (Model and Statistics) methods need to be setup due

to the high demand from the society.

Systematically evaluate the model performance in terms of model

skill and biases. Intensify the efforts to identify causes of biases in

the simulation of thunderstorm in numerical model (e.g., WRF) and

develop/improve the physical processes. The model simulated CG-

lightning flash counts results are compared with Maharashtra

Lightning Detection Network (MLDN) data.

Development towards Lightning simulation

Presently, there is hardly any mechanism which provides a

forecast to issue a warning prior to the occurrence of lightning

and the lightning activity is a typical phenomenon of severe

weather characterized by strong convection.

Conventional approach for thunderstorm prediction using

dynamical model obsevations are well documented and have

some limitations (Mukhopadhyay et al., 2003,2005; Chaudhari et al.,

2010, Ghosh et al., 2004, Rajeevan et al., 2010, Madhulata et al., 2013).

Dynamical Lightning parameterization and dynamical

lightning potential index should be implemented in the

dynamical model along with microphysics for the proper

feedback and coupling .

More and more observed data are now used for the

verification and improving modeling system and further model

development activity.

Dynamical Model set-up

Future research

Objectives: To develop a system for lightning/thunderstorm prediction using dynamical model

Convection Microphysics Lightning para.

M - I Yes Yes (2m & 1m) * Yes – PR92

M - II No Yes (2m & 1m) # Yes – PR94

M - III No Yes (2m & 1m) $ Yes - LPI

* Price, C., and D. Rind (1992) – Based on buoyancy & convective flux,

CAPE, cloud condensate# Price, C., and D. Rind (1994) - Based on dbz, CAPE, cloud condensate$ Yair et al., (2010) – Vertical velocity & cloud condensate

Lightning Parameterization

Lightning Potential Index (LPI)

Radar reflectivityCG-flash counts (#) TS: 27042017

TS: 13052017

Cherrapunje (Meghalaya); East Khasi hills districts of Meghalaya; Bangladesh) [Multiple cell,

max_reflectivity 49 dBz], isolated TS over Kolkata

Thunderstorm occurred over Gangetic WB and Odisha (source: IMD FDP Storm report)

East India region:

Maharashtra region:

TS 27042017-RADARTS 13052017-RADAR

Southern Peninsula (SP) region:

Rainfall

Event MP-06 MP-10 MP-16 MP-17 MP-18

29042017 GPM 0.251 0.294 0.257 0.301 0.305

TRMM 0.38 0.445 0.415 0.399 0.392

05052016 GPM 0.404 0.409 0.398 0.402 0.410

TRMM 0.437 0.468 0.451 0.476 0.501

15032017 GPM 0.313 0.318 0.321 0.318 0.311

TRMM 0.361 0.368 0.376 0.377 0.332

15 TS cases CG lightning flash

counts

Rain (TRMM 3B42) Rain (GPM)

Correlation

Coefficient

0.65

(varies: 0.55 – 0.75)

0.41

(varies: 0.35 – 0.55)

0.32

(varies: 0.25 – 0.45)

Events PSS TS ETS HR FAR SR

TS (15) 0.65 0.51 0.39 0.85 0.19 0.56

Skill Score

Verification & validation

Understanding cloud processes:

Das, S. K. et al., (2019)

Problem identified (DSD)

Strongly Electrified Weakly Electrified

Mudier et al., (2019)

Role of aerosols in severe storm (e.g., TS)

Lightning/Radar data assimilation.

Improvement of model physical parameterization.

New ‘Electric filed’ parameterization along with cloud-

aerosol interaction.

More lightning/Radar data all over the India are required.

Statistical method (e.g., PCA/LDA technique based on Ghosh

et al, 2004; Rajeevan et al., 2010) need to be tested.

First time in India, New approaches for dynamical

‘lightning parameterization’ or ‘lightning potential’

schemes are introduced in the dynamical model (e.g.,

WRF).

Now the present set-up of the regional climate model

(WRF) can simulate cloud-to-ground (CG) lightning

flashes directly (online).

The dynamical Lightning Potential Index (LPI) also

implemented in WRF first time to simulate thunderstorm.

The results are validated with IITM observed ‘lightning

flash count’ data. Results of Correlation and different

verification skill scores shows hope for lighting/TS now-

casting.

Basic research for understanding physical processes for

thunderstorm are carried out for further improvement.

Achievement

Greeshma, M. et al., (2019)

Gayatri, V. et al., (2019)

Hazra, A. et al., (2019)