Validating Low-cost Air-Quality Sensors against Government sensors
612 views
Skip to first unread message
Nafis
May 21, 2012, 11:32:48 PM5/21/12
to airqua...@googlegroups.com
I've been scarfing the local New York State Dept. of Environmental Conservation Air Monitoring station and putting it into COSM (https://cosm.com/feeds/57363). I'm trying to determine the PM25 transfer function from
my Dylos laser particle sensor (https://cosm.com/feeds/55522) and Grove dust sensor (https://cosm.com/feeds/56469). I have given up on using the Sharp Dust sensor for low particle concentrations... I just see noise and no trends.
Ed have you talked to anyone at DEC? I'm trying to put together a grant for a network of air quality sensors as a school project. I'm thinking of calling the local office ( Bureau of Air Quality Surveillance at (518) 402-8508) and see if they would allow us to co-locate some various low-price/lower quality sensors at their site. The goal would be to identify which sensor can actually show useful trends
It looks like there are several station in NYC Region 2 (http://www.dec.ny.gov/airmon/regionMap.php?cityno=2), maybe you could do the same?
Ed Borden
May 22, 2012, 6:16:08 PM5/22/12
to airqua...@googlegroups.com, Joe Saavedra, Tim Dye
Co-locating sensors is definitely a question for Tim Dye and Joe
Saavedra. I've CC'ed them. I haven't myself spoken to anyone at any
government organizations, but these fellas are right in there.
Saavedra. I've CC'ed them. I haven't myself spoken to anyone at any
government organizations, but these fellas are right in there.
Martin Dittus
May 30, 2012, 6:12:33 AM5/30/12
to airqua...@googlegroups.com
This would be an amazing project. I (and surely many others) would be very interested in the results, once you have them. I assume that'll take a few months though…
Projects like this could help us make choices about sensor components. E.g. if we're still evaluating misc hardware designs/component choices by the time this kicks off we could provide multiple sensor variants and then compare.
Out of curiosity, do you already have a list of the particular sensing devices you plan to acquire?
It would be good to also document the respective sensing components, not just the devices used. (I have a very poor understanding of sensor design, and don't know much about the nature and impact of any auxiliary circuit components, e.g. filters; but I expect they are likely much harder to identify and describe.)
And at the risk of derailing the discussion (though still relevant to the subject line) -- have people seen any comparable studies? Both the experimental protocols used and the study outcomes could be relevant to our work here.
m.
Projects like this could help us make choices about sensor components. E.g. if we're still evaluating misc hardware designs/component choices by the time this kicks off we could provide multiple sensor variants and then compare.
Out of curiosity, do you already have a list of the particular sensing devices you plan to acquire?
It would be good to also document the respective sensing components, not just the devices used. (I have a very poor understanding of sensor design, and don't know much about the nature and impact of any auxiliary circuit components, e.g. filters; but I expect they are likely much harder to identify and describe.)
And at the risk of derailing the discussion (though still relevant to the subject line) -- have people seen any comparable studies? Both the experimental protocols used and the study outcomes could be relevant to our work here.
m.
David Holstius
May 30, 2012, 1:35:08 PM5/30/12
to airqualityegg
Chris, I want to thank you for your posts on the Dylos, Sharp, and
Shinyei sensors. I actually went out and bought a bunch of the
Shinyeis as a result, and I've been getting correlations between the
Dylos and the Shinyei PPD42NS. I posted a bit about it here, which
other folks might find interesting:
<http://www.davidholstius.com/2012/05/30/shinyei-vs-dylos/>
The feeds are public (see links on that page).
I'd be very interested in talking more about the school project grant
idea! (maybe offline?) We've done a little bit of workshopping with HS
students here in San Francisco. I'm in touch with a few folks at CA
Dept of Public health who are into the idea of community AQ monitoring
networks. I wonder what it'd look like to do an East Coast/West Coast
sister-schools project.
Re: PM2.5: Data-adaptive methods should in principle be OK for
learning the transfer function, so that's actually something we're
looking into as well. We're very interested in the same topic here at
my lab (UC Berkeley School of Public Health). The biggest
consideration (so the aerosol experts tell me) is that a PM2.5
transfer function is going to be locale-specific, dependent on the
relative size distribution for typical aerosols in that area, as a
function of the predominant sources. So, PM that's geological in
origin (ex: sea salt, resuspended dirt) is going to have different
properties than PM that's anthropogenic (ex: combustion). And it looks
very different in different regions across the United States, at
least.
Roger Unger (of Dylos fame) is interested in a PM2.5 transfer function
too, for basically the same reasons: mass concentrations carry
scientific (as well as political) weight, whereas counts don't (though
that's may change as more research develops around ultrafine PM). We
should get him on this list. But anyway, the overall point is that
there's going to have to be a protocol for learning the transfer
function, rather than a universal transfer function ... and it may
have a lot to do with co-location with authoritative monitoring
stations.
Shinyei sensors. I actually went out and bought a bunch of the
Shinyeis as a result, and I've been getting correlations between the
Dylos and the Shinyei PPD42NS. I posted a bit about it here, which
other folks might find interesting:
<http://www.davidholstius.com/2012/05/30/shinyei-vs-dylos/>
The feeds are public (see links on that page).
I'd be very interested in talking more about the school project grant
idea! (maybe offline?) We've done a little bit of workshopping with HS
students here in San Francisco. I'm in touch with a few folks at CA
Dept of Public health who are into the idea of community AQ monitoring
networks. I wonder what it'd look like to do an East Coast/West Coast
sister-schools project.
Re: PM2.5: Data-adaptive methods should in principle be OK for
learning the transfer function, so that's actually something we're
looking into as well. We're very interested in the same topic here at
my lab (UC Berkeley School of Public Health). The biggest
consideration (so the aerosol experts tell me) is that a PM2.5
transfer function is going to be locale-specific, dependent on the
relative size distribution for typical aerosols in that area, as a
function of the predominant sources. So, PM that's geological in
origin (ex: sea salt, resuspended dirt) is going to have different
properties than PM that's anthropogenic (ex: combustion). And it looks
very different in different regions across the United States, at
least.
Roger Unger (of Dylos fame) is interested in a PM2.5 transfer function
too, for basically the same reasons: mass concentrations carry
scientific (as well as political) weight, whereas counts don't (though
that's may change as more research develops around ultrafine PM). We
should get him on this list. But anyway, the overall point is that
there's going to have to be a protocol for learning the transfer
function, rather than a universal transfer function ... and it may
have a lot to do with co-location with authoritative monitoring
stations.
Chris Nafis
May 30, 2012, 6:11:13 PM5/30/12
to airqua...@googlegroups.com
I'm glad other folks are getting reasonable trend results with the Shinyei
PPD42NS. I think it will be a reasonable sensor to use with the air egg
quality project. The cost is OK ($15) and it seems to show the same trends
the Dylos does. Actually it would be nice to support the Dylos too if you
can afford it.
Let's try to get as many particle sensors co-located with "official" sensors
as possible. The historical data will help researchers like your self
develop better models.
PPD42NS. I think it will be a reasonable sensor to use with the air egg
quality project. The cost is OK ($15) and it seems to show the same trends
the Dylos does. Actually it would be nice to support the Dylos too if you
can afford it.
Let's try to get as many particle sensors co-located with "official" sensors
as possible. The historical data will help researchers like your self
develop better models.
Message has been deleted
Ed Borden
Jul 11, 2012, 9:43:59 PM7/11/12
to airqua...@googlegroups.com
I totally forgot about this thread and re-found it referenced in Martin's "Data Quality" page on the wiki. Chris, did you ever get anywhere with this? We really should be able to at least talk to someone about this in NYC. Co-location has also already been brought up by some of the Chicago participants in an upcoming Egg workshop there (
http://www.meetup.com/sensemakers/events/72719592/ ), and it looks like there are some in's to get it done.
On Thursday, May 31, 2012 4:13:44 PM UTC-4, Karoline Johnson wrote:
On Thursday, May 31, 2012 4:13:44 PM UTC-4, Karoline Johnson wrote:
Hello,
I am researching low cost air quality monitors for the US EPA. I will
be testing them against our monitors over the next few months. You
have mentioned the Shinyei, Dylos, and the Grove dust sensors but if
you have any other suggestions of low cost air quality monitors (PM or
others) to test I would be very interested to hear your suggestions.
Thanks,
Karoline
> > the same?- Hide quoted text -
>
> - Show quoted text -
Chris Lavenson
May 7, 2014, 11:06:11 AM5/7/14
to airqua...@googlegroups.com
On Thursday, May 31, 2012 4:13:44 PM UTC-4, Karoline Johnson wrote:
> Hello,
>
> I am researching low cost air quality monitors for the US EPA. I will
>
> be testing them against our monitors over the next few months. You
>
> have mentioned the Shinyei, Dylos, and the Grove dust sensors but if
>
> you have any other suggestions of low cost air quality monitors (PM or
>
> others) to test I would be very interested to hear your suggestions.
>
> Thanks,
>
> Karoline
>
>
>
> On May 30, 6:11 pm, "Chris Nafis" <na...@nycap.rr.com> wrote:
>
> Hello,
>
> I am researching low cost air quality monitors for the US EPA. I will
>
> be testing them against our monitors over the next few months. You
>
> have mentioned the Shinyei, Dylos, and the Grove dust sensors but if
>
> you have any other suggestions of low cost air quality monitors (PM or
>
> others) to test I would be very interested to hear your suggestions.
>
> Thanks,
>
> Karoline
>
>
>
> On May 30, 6:11 pm, "Chris Nafis" <na...@nycap.rr.com> wrote:
>
> > I'm glad other folks are getting reasonable trend results with the Shinyei
>
> > PPD42NS. I think it will be a reasonable sensor to use with the air egg
>
> > quality project. The cost is OK ($15) and it seems to show the same trends
>
> > the Dylos does. Actually it would be nice to support the Dylos too if you
>
> > can afford it.
>
> >
>
> > Let's try to get as many particle sensors co-located with "official" sensors
>
> > as possible. The historical data will help researchers like your self
>
> > develop better models.
>
> >
>
> >
>
> >
>
> > ----- Original Message -----
>
>
> > PPD42NS. I think it will be a reasonable sensor to use with the air egg
>
> > quality project. The cost is OK ($15) and it seems to show the same trends
>
> > the Dylos does. Actually it would be nice to support the Dylos too if you
>
> > can afford it.
>
> >
>
> > Let's try to get as many particle sensors co-located with "official" sensors
>
> > as possible. The historical data will help researchers like your self
>
> > develop better models.
>
> >
>
> >
>
> >
>
> > ----- Original Message -----
>
> > > the same?- Hide quoted text -
>
> >
>
> > - Show quoted text -
Karoline:
>
> >
>
> > - Show quoted text -
I am very interested in finding your results from the study you referenced above. Do you have some time we can talk? I am working on a project for our city and would love to see what you thought were the most accurate low cost solutions. please email me at clav...@durhamlabs.com
Reply all
Reply to author
Forward
0 new messages