Air Particle Sensor in Doha, Qatar

Introduction

One of the downsides of living here in Qatar is the dusty air, but there's very little data available, and there are no stations that track the air quality in real time and post the data to the internet. So shortly after I accepted an offer to move here in February 2013, I began researching a way of tracking the air quality myself and posting my findings to the internet.

Had it not been for an engineer named Chris Nafis sharing his work on the subject, I probably would've given up a long time ago. But I stuck with it, and I now have two air particle sensors here in Doha which post air particle counts to the internet every 30 seconds.

I put my first air particle sensor in our living room to get a sense for the air quality we were breathing on a daily basis. Before long, I realized that the data was heavily affected by things like cooking and vacuuming and didn't give a very accurate picture of what the air quality in Doha is really like. So I decided to build another sensor and put it outside.

Here's a picture of the basic setup I'm now using outside:

Arduino Ethernet attached to particle sensor sitting on my windowsill outside

I bought the plastic container at the supermarket and drilled holes to allow air to pass through. Inside is the particle sensor and an Arduino Ethernet which manages the data crunching and posting to the internet. There's a screen at the top which shows hourly averaged values as well as the temperature and humidity. All of this sits on the front, north-facing windowsill of our house on the first floor.

Here's a live graph of the outside data from the past three days:

graph

If you'd like to go directly to the data and see what the air in Doha is like right now, click here.

What is air particulate?

Air particulate is a type of pollution. It can be made up of pollen, smog, smoke, dust, and other things. The more particulate there is in the air, the more health problems people have (though there doesn't seem to be a consensus on how particulate affects health and to what degree), not to mention its effect on visibility and quality of life. Wikipedia has a more complete discussion of the impacts of air particle pollution.

How is it measured?

There are two ways to measure how much particulate is in the air: you can weigh it (by filtering a known volume of air through a filter and then weighing the filter), or you can count individual particles using a laser. The former is more exact, but it's also much more expensive, so DIY-types like myself use the cheaper laser counters. Unfortunately, the data from the one I use is very noisy (as in, it'll read 200 particles per 0.01 cubic foot one minute, and then 800 the next). But when the values are averaged out, the trend lines are more clear.

The other challenge with using these cheap laser counters is that the data they output is in different units (number of particles of a certain size in a certain volume of air) than what professional air quality measuring stations use (weight of combined particles in a certain volume of air).

A note on units: The particle sensor I use measures the number of particles within a volume of air equal to 0.01 cubic feet, or 283 milliliters. This volume is close to the same volume as a baseball, or the tidal volume (regular inhalation) of a 10 year-old child weighing 70 pounds.

The upshot of all of this is that it's difficult to compare my data with the data in other places around the world. However, a guy based in Beijing named Thomas Talhelm has attempted to correlate his laser particle counter (a Dylos 1100) with professional governmental figures. Still, my data should be taken with a big grain of salt for reasons I explain below.

Disclaimer

Just in case it's not already apparent, this is very much an amateur operation using amateur equipment. I am not a professional scientist, nor do I have any formal training in meteorology, pollution, or electrical engineering. I'm just interested in this stuff and have followed my own curiosity.

That said, I believe that my outdoor dust sensor gives a decent relative look at the air quality in Doha. Is it dustier here today than it was yesterday, or last week, or a month ago? Check the dust sensor data to get a good idea. But can it tell you definitively how the air here compares to the air in New Delhi or Beijing or New York? Probably not.

Equipment

I use a Shinyei PPD42NS particle sensor connected to an Arduino Ethernet. The data is pretty noisy, but taking averages smoothes it out. If you'd like to see a complete guide of how to build an air particle sensor and connect it to the internet, please see my guide on Instructables.

Findings

Before I get into the information I've been able to gather from my air particle sensor, I should point out that the setup seems to work well. My indoor sensor has been running without issue for a year and a half and my outdoor sensor has been running for nearly a year. I haven't had to restart either of them or do anything else to them at all. They just plug away, posting data to the internet. Both sensors occasionally post zero values (a value which indicates that there is no particulate in the air whatsoever) which I take to be errors, as I'm sure there's always a little bit of particulate floating around. These zero values are generally confined to times when the particle sensors are saying that the air is relatively clean (almost never when it says the air is really dirty), so I assume that this is just a product of having relatively clean air. Still, I'd like to figure out what exactly is causing these zero values and eradicate them.

Annotated air particulate chart

The chart above shows a three day span of data from my indoor particle sensor. You can clearly see three events which cause the amount of particulate in the air to spike. The first is a minor dust storm, the second is when I cooked dinner, and the third is another dust storm that happened after we went to bed. I don't know for sure, but it seems reasonable to assume that wind blows dust into the house through imperfect sealing at the doors, windows, and elsewhere, or just when we go in or out of the house.

Here's what the inside of our front door looked like the morning after (note the streams of dust blown in from under the door jam:

Dust blows in under the door

The chart illustrates two things I've learned since I set up the air particle sensor:

  • Poor air quality outside (like what you see from a dust storm) has a clear and measurable effect on inside air quality.
  • Frying vegetables in a cast iron pan puts a ton of particulate into the air, and it spreads throughout the house (the particle sensor is in another room and about 13 meters from our stove).

I haven't put the numbers into Excel yet, but it looks like the number of particles in the air follows a sort of exponential decay after we pollute the air by frying. The reversion to "normal" values after the wind here kicks up a lot of dust happens more slowly, but it's still noticeable.

Does Doha have bad air?

It seems pretty clear that air quality in Doha isn't very good (buildings that are relatively nearby are often obscured by haze, and at night, you can see the beams of car headlights), but without quantification it's hard to say how bad it is, and how it compares to air quality in other places. In 2011, the government of Qatar published data which showed that average air particulate matter far exceeded the limit of 50 micrograms per cubic meter (which I believe they pulled from the European Environment Agency). Unfortunately, the original report doesn't go into much detail about how the data was gathered, and I've yet to find a more current report.

According to the 2011 government report, PM10 (air particulate between 2.5 and 10 micrometers in diameter, more info here) in Doha ranged from 105 micrograms per cubic meter to 185 micrograms per cubic meter. The closest location to me is the Corniche, where values averaged 120 micrograms per cubic meter. This corresponds to an Air Quality Index (AQI) value of 83 ("Moderate"). This allows comparison with values all over the world. For example, the AQI for PM10 right now (March 15, 2014) in my previous home of Portland, Oregon, is 14 ("Good"). In Beijing right now, the PM10 AQI reading is 104 ("Unhealthy for sensitive groups"). So Doha appears to be somewhere in the middle between Beijing and Oregon, but this is a very, very rough estimate.

DohaDust.org

I like to look at a graph to see how the air in Doha is changing over time, but I figured that some people might just want to get a quick snapshot, so I built DohaDust.org which pulls an hourly average of the data from the outdoor sensor (via Xively) and gives it a qualitative rating relative to an average day in Doha ranging from "excellent" to "good" to "fair" to "poor."

There are three major caveats about how I've designed DohaDust.org:

The first is that the ratings are simply qualitative descriptors that I've come up with after reviewing six months of air quality data in Doha and comparing them with how I would've described them to someone asking "How's the air in Doha right now?" Here's how those descriptors match up with the average, hourly value being put out by the outdoor dust sensor:

  • Excellent: under 300 particles per 0.01 cubic feet
  • Good: 301 to 700 particles per 0.01 cubic feet
  • Fair: 701 to 1200 particles per 0.01 cubic feet
  • Poor: 1200 particles per 0.01 cubic feet and above
Basically, an "excellent" reading corresponds with when it's really, really clear outside and a "poor" reading corresponds with a dust storm. As I wrote above, 0.01 cubic feet is close to the volume of a baseball or the amount a 10 year-old, 70 pound child breathes in a single breath.

The second is caveat is that this method of describing the air quality in Doha is meant only as a comparison to Doha itself. If I were comparing the air in Doha to the air in Portland, Oregon, I'm sure it'd never be anything besides fair and poor. But that's not very helpful, so I've stuck with a wider range of descriptors.

The final caveat that bears mentioning is that my project only measures air particulate. The sensor has nothing to say about sulfur oxides, or nitrogen oxides, heavy metals, or any other form or air pollution.

To do list

This is an ongoing project. Here are a few things that I'd like to improve in the future: