Should IMD use Weighted Averages for Evaluating Southwest Monsoon?

The performance of Southwest Monsoon is evaluated by IMD through a weighted average method under which the sub divisional rainfall is averaged to be tallied for a national average rainfall.  Does this provide an accurate picture of how the rainfall has been for the season? A COMK alternate take on the performance of Southwest Monsoon.  The study has been featured in Business Line today thanks to Mr. Vinson Kurian who also thinks IMD should evolve with time in how rainfall numbers are measured.

Current status of Southwest Monsoon (As of 28th July 2015 along with couple of other milestone points used for reference in this study)

SWM1

As one can see while overall the country is still in the Normal Zone the average rainfall until 28th July is -3% which is pretty much par for course.

So Is India having near normal rainfall, how true is it? Now let us do a little bit of number crunching on this. For this we are taking up the best performing region so far Northwest India.

SWM2

In the table above we have featured North West India which has the highest positive anomaly among all regions at 12% higher than normal rainfall.  Even in this region J&K, West Rajasthan & East Rajasthan are the only positive Sub Divisions with all others having Negative Anomaly though except for East UP all the other divisions are in the normal category because of the -20% to 20% classification.

SWM3

Now let us link the same with how the rainfall is spread out by using the individual district status.  In this it is quite clear while the region is overall 12% more than normal rainfall, 38% of the area is still under either Deficit or Worse condition.  The obvious pitfall of the average system shows up in the composite Uttar Pradesh state where West UP has more than half of the state is in deficit but the overall average is only 12% lesser than normal.  When one compares West Uttar Pradesh & Haryana, Chandigarh & Delhi is compared both have an almost equal -12% anomaly cumulatively as of 28th July but while West UP has 53% of the districts in deficit or worse category comparatively Har,CHD& Delhi has only 35% of the districts in deficit category.

Now let us look at the worst performing region, Southern Peninsula to check if the weighted average method indicates the situation on the ground well.

SWM4

As one can observe the entire region has 50% of the districts under either Deficit or Worse categories with only South Interior Karnataka & Coastal Andhra Pradesh having a less spread of deficit districts.  The situation is extremely alarming in the regions of Kerala, North Interior Karnataka & Rayalseema where at least 75% of the districts are in deficit category. But under weighted average the region is still under normal category

So how does the National picture looks like does the weighted average method showcases the spatial variations in the rainfall pattern

SWM5

It is quite clear that even when Southwest Monsoon was at 16% nationally which puts it under excess category only 70% of the country had received normal or better rains.  As things stand only about 62% of the country has received normal rainfall cumulatively this season as of 28th July.

Also if one keenly observes in the case of East & North East India and North West India while the weighted average method shows negative progress between 30th June & 26th July to the contrary the number of districts with normal or better rainfall increases in both the regions confirming the rains have geographically better spread during this period which is missed out by the Weighted Average method.

Finally when one looks at an individual day’s rainfall the flaws of weighted average method becomes glaringly obvious.  28th July the National average was 32% in excess of the normal rainfall unfortunately when one goes through the district break up only 22% of the districts nationally got Normal or Better Rainfall with almost 75% of the country receiving Scanty or worse (-60% Anomaly), a few districts receiving unprecedented extremely heavy rainfall is skewing the national average providing a far from accurate picture.

SWM6

Is it time to create a new way of providing the progress of Southwest Monsoon & judging whether the country has been receiving average rainfall or not. When excess rainfall in Rajasthan cannot provide any comfort to the deficit state of Kerala why the weighted average system should be adopted.  Instead should we not confirm how much of the area have received normal or better rainfall cumulatively providing a much more accurate representation of the spatial spread of Monsoon.

A complete summary of Southwest Monsoon so far, across all divisions a very detailed table that gives the birds eye view of the season so far.

SWM_Summary