March 20  2005 Tornado in Northern Bangladesh
 
 
 

        A tornado struck northern Bangladesh on March 20, 2005, killing 50 people.  The tornado affected area
        was centered near 89.6E and 25.4N.

        I will attempt here to give those interested souls an idea of the synoptic setup for this event. Unfortunately
        detailed information about the low-level shear profile is not available. But the wind profile from 700 to
        200mb can be approximated, along with the surface conditions. Thus, I can provide a rough estimate of
        instability and shear. Unfortunately the quality of the Indian radiosonde data is very low.

        The elevated mixed layer is not yet well developed across the Bengal region. But India has been heating up
        and slowly drying out over the past few weeks. This process will be maturing in early to mid April when well
        defined drylines will be a daily occurrence. Elevated mixed layers in March tend to be more shallow in nature,
        but become very deep--as high as 550mb later in April.

        Because of the terrain, the mid-level winds tend to be enhanced immediately south of the Tibetan Plateau.
        So the 700-500mb winds are greatly enhanced by flow around the Plateau. This is because the average
        elevation is almost 500mb. Actually the 400-300mb flow also gets deverted south of the Plateau. It does not
        suffice to have the strong mid-level winds only. The high level jet needs to be in a favorable position relative
        to Bengal for high-end severe storms to  occur. The high level jet both aids in large scale lift and evacuate
        precip from the updrafts, leading to longer lived storms. So the high level jet is crucial to the development
        of long lived supercells.

        Oftentimes, the 250mb jet is located along the extreme southern edge of the Tibetan Plateau, with the
        mid-level jet  to the south over central and northern Bangladesh. In such cases, a massive CAPE robber effect
        can occur with 300mb temps warming to perhaps -28C. Sometimes even the mid-level flow retreats north of the
        Plateau, but usually there is at least a trickle of flow south of the Plateau even into late May.

        In this tornado case, the high level jet extended from North Africa across southern Iran and then southeast
        across the Bengal region(UKMET 250mb initialization from 12 UTC March 20). 400-300mb temps were fairly
        cool in the left exit region of the high-level jet. 500mb temps were fairly cool ~11C.  The 12 UTC surface
        chart showed dewpoints in the lower mid 70s(22-24C) with surface temps in the lower to mid 80sF(25-28C).
        This yields lifted indices around -10 which of  course translates to high instability.  400 and 300mb lifted indices
        were around -13C. Note the 700mb jet south of the Tibetan Plateau over northern Bangladesh(40kts at 00 UTC
        increasing to 45 to 50kts by 18 UTC). 500mb winds were between 40 and 50 kts. This is actually very typical in
        tornadic situations--mid level winds 40-50kts.

        Maps for your amusement:

        Surface 12 UTC March 20  tornado location denoted as T
        500mb 12 UTC March 19
        500mb 00 UTC March 20
        500mb 12 UTC March 20
        500mb 00 UTC March 21
        250mb reanalysis 
        500mb UKMET 00 hr valid 06 UTC March 20
        500mb UKMET 06 hr valid 06 UTC March 20    heights/winds/temprature
        500mb UKMET 06 hr valid 06 UTC March 20    heights/winds/temperature/vorticity
        500mb UKMET 00 hr valid 12 UTC March 20    heights/winds/temperature
        500mb UKMET 00 hr valid 12 UTC March 20    heights/winds/temperature/vorticity
        500mb UKMET 00 hr valid 12 UTC March 20    heights/winds/temperature--small view
        500mb UKMET 06 hr valid 18 UTC March 20    heights/winds/temperature--small view and location of tornado(T) shown (15 UTC)
        700mb UKMET 00 hr valid 12 UTC March 20    winds/temps--very large view
        700mb UKMET 00 hr valid 12 UTC March 20    winds/temps--large view
        700mb UKMET 00 hr valid 12 UTC March 20    winds/temps--small view
        700mb UKMET 06 hr valid 18 UTC March 20    winds/temps--large view and location of tornado(T)
        700mb UKMET 06 hr valid 18 UTC March 20    winds/temps--small view
        250mb UKMET 06 hr valid 06 UTC March 20    250mb windspeed/winds--large view
        250mb UKMET 06 hr valid 06 UTC March 20    250mb windspeed/winds--small view
        250mb UKMET 06 hr valid 06 UTC March 20    250mb windspeed/winds--small view
        250mb UKMET 06 hr valid 06 UTC March 20    250mb windspeed/winds--large view
        300mb UKMET 06 hr valid 06 UTC March 20    300 windspeed/winds/temps/large view
        300mb UKMET 06 hr valid 06 UTC March 20    300 windspeed/winds/temps/small view
        300mb UKMET 00 hr valid 12 UTC March 20    300 windspeed/winds/temps/small view
        300mb UKMET 00 hr valid 12 UTC March 20    300 windspeed/winds/temps/large view

        I do not have access to satellite data for this event. But surface observations indicate thunderstorm activity
        across northern Bangladesh the night before and the morning before the tornado. Thus, there was likely an
        outflow boundary in the region of interest. This would allow for backed surface winds, low LCL heights and
        high instability immediately on the cool side of the boundary.