North-east update

A new weather setting has taken hold compared to that of early last week. Then, an upper trough straddled northern India and southwesterlies (the last of the South West Monsoon) were still wetting the far-northeast of the Subcontinent. Now, high pressure has built aloft over the Subcontinent and has shut off the SW Monsoon over Bangladesh and nearby India. Farther south, a broad low has spread over southern India and Sri Lanka, the results being the blow-up of showers and thunderstorms suggested by the above infrared cloud shot (IMD).

The pressure as of Monday evening, local time, is 2-4 millibars higher over Bangladesh than over Sri Lanka. This is enough to drive a weak (north)easterly wind flow over the western Bay of Bengal, so we should be looking at the first days of a North East Monsoon.

Numerical forecasts do not agree as to the coming weather for the greater North Indian Ocean region, although there are important aspect wherein there is agreement. As I had talked up last week, there is still the suggestion that a tropical depression (or cyclone) will happen over the Bay of Bengal later this week. Timing for a visit to the Tamil Nadu coast of India was indicated for the 26th or the 27th. Today, numerical models are not agreed as to this (other than the NOGAPS, which still shows it).

One cause of the straying of the model solutions lies with the low over southern India. There is an tendency on the part of the numerical forecasts to take this low to the southern Arabian Sea, where it would compete with and perhaps slow the growth of any Bay low. By the way, the seeds of any Bay low may already be sown in the blow-up of cloud and rain over Sumatra and the Malay Peninsula (see the satellite image). The last two to three days have seen rainfall of 33 cms, or 13.0 inches, at Sibolga, western Sumatra.

Both the GFS and the ECMWF show lows over the Arabian Sea looking forward 5-7 days. And the longer range outlook of the GFS even shows low pressure reaching southeastern Arabia with rain. This is forecast about two weeks forward in time--far too far ahead for any meaningful confidence.