Measuring loudness on Deepavali nights

The Indian festival of Deepavali gets its name from the Sanksrit for “display of lights”, “Deepaanaam aavali“. These days, the festival is anything but about lights, especially in urban centers where the bursting of loud firecrackers has replaced the gentler display of lamps. Sometimes, Bangalore – where I live – sounds like a warzone. People I’ve spoken to have defended the way they celebrated it, saying, “It’s a tradition thousands of years old!” No, it’s not.

The Noise Pollution (Regulation and Control) Rules 2000 (“Rules”) limits the amount of environmental noise by area and time of day. In residential areas, for example, the maximum allowable noise level between 6 am and 10 pm (‘daytime’) is 55 dB(A) Leq.

A Central Pollution Control Board document released on October 24 reports the results of an exercise where noise-level monitors listened in in five areas of the national capital, Delhi, for the week leading up to Deepavali: October 15 to October 23. Without exception, the dB(A) Leq readings in all five areas – Pragati Maidan, East Arjun Nagar, NSIT Dwarka, IHBAS Dilshad Garden and DCE Bawana – have increased from 2013 to 2014. The nighttime readings breach the Rules limits by at least 10 dB(A), which warrants a complaint.

Insofar as the Rules is concerned, the units of measurement play a defining role in how meaningful the limits are.

For starters, dB stands for decibels, a logarithmic measure of noise levels. According to ISO standards, a doubling of noise levels is equal to an increase of 3 dB.

Because noise levels during many kinds of measurements – including during Deepavali – keep changing, Leq is used because it denotes an average noise level during a specified period. Moreover, because dB is a logarithmic measure, Leq cannot be calculated like a simple average. Instead, sound-meters usually convert dB into the corresponding sound pressure levels and then calculate the average. In the process, the A-setting is also applied: it is a scale to measure the perceived human loudness.

As it is, the Rules don’t explicitly specify the time period across which the noise levels are to be measured. The only mention of periods, in fact, is when the document defines daytime (6 am to 10 pm) and nighttime (10 pm to 6 am). So the noise level of 55 dB(A) Leq is presumably defined for a 16-hour period (daytime; residential area). An obvious outcome of this is that infrequent loud noises in a generally quiet residential area will not breach the legal limits during daytime.

But what about during Deepavali? Let’s say the festival is being celebrated on a weekday: the bursting of firecrackers will start around 4 pm (once the kids have returned from school) and last until 9 pm. Could noise levels in this five-hour period push the daytime average beyond 55 dB(A) Leq?

I used the NoiseTube project’s mobile app (of the same name) that makes per-second measurements and calculates the minimum, maximum and average dB(A) Leq over a specified duration. Sitting about 80 m from a bunch of kids bursting firecrackers in our apartment driveway, I used the app to make 300 measurements over 5 minutes for the following results:

Min.: 41.51 dB(A) Leq
Max.: 83.88 dB(A)Leq
Avg.: 66.41 dB(A) Leq

Earlier in the day, I’d made a five-minute measurement when no firecrackers were being burst for an avg. reading of 42 dB(A) Leq. So, assuming that 42 dB(A) Leq was the reading for 11 hours and 66.41 dB(A) Leq for the remaining five, the daytime average reading comes to 56.53 dB(A) Leq. Abiding by the Rules, this isn’t even enough for me to register a complaint, which necessitates the noise levels to exceed the limit by at least 10 dB(A).

At the same time, the noise levels are debilitating. When the cracker-bursting frenzy is in full swing, I’ve recorded noises louder than 100 dB(A). If I spend a day outside, and if the sulfurous fumes don’t give me a migraine, just the noise will.

Because of this, the Rules might be more meaningful – and effective – if a measurement duration is defined, such as between certain times of day according to what time of the year it is (correct me if I’m wrong because I’d love to be wrong about this). In fact, because dB is logarithmic, any average will be biased toward the higher values (as exponentially higher the numbers, higher the logarithms), and even with this boost, 66.41 dB(A) Leq over five hours is not ‘illegal enough’.

Featured image from Wikimedia Commons