Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

New particle formation event #602

Closed
markusstocker opened this Issue Apr 1, 2018 · 25 comments

Comments

@markusstocker
Copy link

markusstocker commented Apr 1, 2018

I had a meeting with the research community in Helsinki, including @mazaidan. It was pointed out that "formation of particles in an atmosphere" really should be called "new particle formation event". The researchers felt very strongly about having new and event in the concept. It was also pointed out that a process is not observed, rather the event is observed. To my understanding, the event consists of many processes. The point is, the correct naming of these terms isn't settled and I like to open this issue to try getting this straight, and I like to directly involve the research community (also Pauli et al. but not sure if they are on GitHub, @mazaidan can you check?).

@cmungall

This comment has been minimized.

Copy link
Member

cmungall commented Apr 1, 2018

we should name the classes using terms used by the research community.

So far ENVO hasn't really recognized any distinction between events and processes at the ontological level. A process is simply something that happens. Connections between smaller units and larger units would be achieved using relations like part-of. But we'd be open to other ways of modeling, especially if this fits better with other ways of modeling. We also try to represent the underlying process rather than it's observation.

A crucial point for the ontology is the level of granularity we're talking about. Does a single instance of particle formation generate a single particle? Or are we talking about assemblages of these? We have somewhat fudged this in analogous biological process ontologies, with the formal definition of metabolic or signaling processes being in terms of individual molecules, but with the classes used to represent statistical aggregates of these.

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented Apr 2, 2018

An instance of a new particle formation event generates a polydisperse aerosol, so multiple particles of different diameter sizes. During the event (which lasts several hours), the aerosol changes, overall the particles get bigger in size. What scientists describe are events, e.g. the duration, particle growth rate. I fully agree we need to model this how the research community conceptualizes. I hope to get some involved here.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented Apr 19, 2018

I would prefer to have separately "new particle formation", where one can describe the process, which can be clearly defined. The "new particle formation event" is not only an occasion of a formation of particles, but it stands for a period of regionally and continuously (e.g. close to or over an hour) elevated atmospheric new particle formation rate. And the threshold for the formation rate to be classified as an event is not very exact, it is basically our capability of detecting elevated rate out of noise.

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented Apr 24, 2018

Thanks @PauliPaasonen and @mazaidan for joining in and your comment. Are you suggesting to have two separate concepts, one for process and one for event?

To my understanding of your (particle size distribution) data analysis workflow, you do describe individual new particle formation events, those occasions of a formation of particles. This is what I understand as "secondary data" generated in analysis (also visually) of particle size distribution data (the "primary data", in this context). The attempt here with ENVO is to provide a "guideline" for how to describe such events, their type, spatio-temporal location, and other attributes such as formation rate. I guess there is uncertainty in all these values (strictly speaking even in spatio-temporal location but also the assessed type (Ia, Ib, II, etc.)). This is another matter, and we may need to think how to capture uncertainty. First, though, we need to develop a scheme to capture the actual values in a coherent data object. Agreed?

From your comment I am unclear how the event relates to the process. I think this needs to be clarified but I wonder if it can be done in a second step?

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented Apr 24, 2018

I think that would be good, to have 2 separate entries. Wasn't it possible to have one under another? My point with the definition of the term new particle formation event (NPF event) is that it stands for regional atmospheric burst of new particles. E.g. new particle formation in the tale-pipe or exhaust plume of a car or industry stack is not a NPF event, at least I have never heard NPF event referring to these kinds of situations of new particle formation.

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented Apr 24, 2018

Excellent. I conclude that we need a concept named "New Particle Formation Event" and I understand this is a kind of New Particle Formation (a subclass), one that does not include things like new particle formation in tale-pipe or exhaust plume.

Now, ENVO currently includes the concept "formation of particles in an atmosphere" with two special kinds, "formation of liquid droplets" and "formation of solid particles", and more specific kinds, e.g. "formation of solid particles from gaseous material in an atmosphere".

@PauliPaasonen: is it accurate to say that a "New Particle Formation Event" (in the context discussed here) is a "formation of solid particles from gaseous material in an atmosphere"? If positive, is "New Particle Formation Event" equivalent to "formation of solid particles from gaseous material in an atmosphere" or, rather, a more specific type (subclass)?

Depending on the answers, @pbuttigieg can we adopt "New Particle Formation Event" into ENVO? This implies a classification of the Event as a Process. I believe there is disagreement on this matter (in the literature). Some consider Event and Process as ontologically distinct. But I leave this to @pbuttigieg and @cmungall and their understanding of ENVO and its alignment with BFO.

To my understanding, for the science community it is important to have the term "New Particle Formation Event" appearing in the ontology, because this is how these events a called. If we are lucky, we can simply align "New Particle Formation Event" to the current class hierarchy.

Possible subclassing of "formation of solid particles from gaseous material in an atmosphere" relies on the correctness of the current description, namely that a "New Particle Formation Event" is a formation of particles that occurs in some atmosphere and has solid aerosol as output (but I suppose this should be correct).

We should then be able to further constrain the concept "New Particle Formation Event".

@cmungall

This comment has been minimized.

Copy link
Member

cmungall commented Apr 24, 2018

Event vs process naming: I'm happy to use whatever terminology the community uses.

Event v process ontologically: BFO doesn't have a distinction between event and process. This doesn't mean we can't introduce logical distinctions where they are important. So far I haven't seen a use case, simplistic treatment of everything as processes in the bfo sense seems to work. Ontologically there are not many commitments here. Each instance of a process has a start and an end (which may be fiat, and may be unknown to us). A process has participants. In our design patterns these participants can be subdivided into inputs, outputs. To me its more important to fit each process/event into a design pattern that to worry about upper level classification.

@dr-shorthair

This comment has been minimized.

Copy link

dr-shorthair commented Apr 25, 2018

Agree with @cmungall here, though I find the label 'Process' a bit non-intuitive because of the English verb with the same name. I try to remind myself to filter it to 'the BFO meaning of Process'. By now this is history and isn't going to change.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented Apr 25, 2018

@markusstocker: "New Particle Formation Event" cannot be only "formation of solid particles from gaseous material in an atmosphere". But "new particle formation", if we want to put it in very simple way, can be "formation of solid or liquid particles from gas-phase molecules". To be more exact, it could be "process, in which gas-phase molecules form solid or liquid particles, which are stable enough to continue growing by condensation of other gas-phase molecules more probably than to evaporate back to gas-phase". "New particle formation event" is an atmospheric phenomenon, in which the new particle formation (from well dispersed gas-phase molecules) is intense enough in large enough area and for long enough time for us to observe the formed particle population. This non-exact definition comes from the history and is related to the detection limits of the instruments and stationary nature of the measurements sites. This would need more time to be well explained. Anyhow, there is no exact thresholds for the intensity, area and duration of the enhanced new particle formation, which could be used for determining whether there is a new particle formation event or not. If elevated new particle formation rate is observed in the atmosphere (away from direct sources of the related gas-phase molecules) for a period of time, it can be called an event. Note that at least the last statement is not objective, even though I tried to make it such.

pbuttigieg added a commit that referenced this issue Apr 27, 2018

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented Apr 27, 2018

Many thanks all for the great discussion here!

In response to #602 (comment)

I had a meeting with the research community in Helsinki, including @mazaidan. It was pointed out that "formation of particles in an atmosphere" really should be called "new particle formation event". The researchers felt very strongly about having new and event in the concept.

We can move the existing label to an alternative label field to preserve common usage. As long as the definition is the same the semantics work. If there are any other synonyms, please let us know. See below for the specifics of how we suggest handling this case.

It was also pointed out that a process is not observed, rather the event is observed. To my understanding, the event consists of many processes.

Again, I don't really see the difference here (as processes can have other processes as parts). As noted above, the idea of this being that part of a process which is observed can be used as a valid differentiae in BFO aligned systems. I do think that the rate differentia suggest by @PauliPaasonen is more clear.

The point is, the correct naming of these terms isn't settled and I like to open this issue to try getting this straight, and I like to directly involve the research community (also Pauli et al. but not sure if they are on GitHub, @mazaidan can you check?).

Great!

From #602 (comment)

An instance of a new particle formation event generates a polydisperse aerosol, so multiple particles of different diameter sizes. During the event (which lasts several hours), the aerosol changes, overall the particles get bigger in size. What scientists describe are events, e.g. the duration, particle growth rate. I fully agree we need to model this how the research community conceptualizes. I hope to get some involved here.

Okay, we can adjust the outputs accordingly and update the aerosol hierarchy.

From #602 (comment)

I would prefer to have separately "new particle formation", where one can describe the process, which can be clearly defined. The "new particle formation event" is not only an occasion of a formation of particles, but it stands for a period of regionally and continuously (e.g. close to or over an hour) elevated atmospheric new particle formation rate. And the threshold for the formation rate to be classified as an event is not very exact, it is basically our capability of detecting elevated rate out of noise.

We could do this by having the event as a subclass of the process differentiated by its increased rate. The semantics to do so exist in PATO and are similar to what we use for storms and other entities that feature processes with increased rates.

From #602 (comment)

Excellent. I conclude that we need a concept named "New Particle Formation Event" and I understand this is a kind of New Particle Formation (a subclass), one that does not include things like new particle formation in tale-pipe or exhaust plume.

I think we're aligned here. We can axiomatise with some of our "natural" differentia to make sure this is understood as a process without a technological source and that occurs in an atmosphere.

Now, ENVO currently includes the concept "formation of particles in an atmosphere" with two special kinds, "formation of liquid droplets" and "formation of solid particles", and more specific kinds, e.g. "formation of solid particles from gaseous material in an atmosphere".

@PauliPaasonen: is it accurate to say that a "New Particle Formation Event" (in the context discussed here) is a "formation of solid particles from gaseous material in an atmosphere"? If positive, is "New Particle Formation Event" equivalent to "formation of solid particles from gaseous material in an atmosphere" or, rather, a more specific type (subclass)?

See below.

Depending on the answers, @pbuttigieg can we adopt "New Particle Formation Event" into ENVO? This implies a classification of the Event as a Process. I believe there is disagreement on this matter (in the literature). Some consider Event and Process as ontologically distinct. But I leave this to @pbuttigieg and @cmungall and their understanding of ENVO and its alignment with BFO.

I think it's much clearer now. In this case, as long as we have a good definition (which was provided by @PauliPaasonen) we don't mind the label text. I'll include an editors' note to indicate that this is the common usage and the label should be retained as discussed here.

Possible subclassing of "formation of solid particles from gaseous material in an atmosphere" relies on the correctness of the current description, namely that a "New Particle Formation Event" is a formation of particles that occurs in some atmosphere and has solid aerosol as output (but I suppose this should be correct).

Yes. See below.

We should then be able to further constrain the concept "New Particle Formation Event".

Yes, much easier now to move forward. As long as we have a good consensus definition, we can worry about handling the upper level stuff while making sure user requirements are met.

I'm with @cmungall in :

Event vs process naming: I'm happy to use whatever terminology the community uses.

As said above, if the def is clear enough, we can make sure it works.

From #602 (comment)

Agree with @cmungall here, though I find the label 'Process' a bit non-intuitive because of the English verb with the same name. I try to remind myself to filter it to 'the BFO meaning of Process'. By now this is history and isn't going to change.

Indeed, but we can insulate the userbase from this and make sure their terminology and synonyms are reproduced faithfully. This will help any NLP or text-mining efforts immensely.

From #602 (comment)

@markusstocker: "New Particle Formation Event" cannot be only "formation of solid particles from gaseous material in an atmosphere". But "new particle formation", if we want to put it in very simple way, can be "formation of solid or liquid particles from gas-phase molecules". To be more exact, it could be "process, in which gas-phase molecules form solid or liquid particles, which are stable enough to continue growing by condensation of other gas-phase molecules more probably than to evaporate back to gas-phase". "New particle formation event" is an atmospheric phenomenon, in which the new particle formation (from well dispersed gas-phase molecules) is intense enough in large enough area and for long enough time for us to observe the formed particle population. This non-exact definition comes from the history and is related to the detection limits of the instruments and stationary nature of the measurements sites. This would need more time to be well explained. Anyhow, there is no exact thresholds for the intensity, area and duration of the enhanced new particle formation, which could be used for determining whether there is a new particle formation event or not. If elevated new particle formation rate is observed in the atmosphere (away from direct sources of the related gas-phase molecules) for a period of time, it can be called an event. Note that at least the last statement is not objective, even though I tried to make it such.

I suggest that we include a comment on the class indicating these concerns, but avoid over axiomatising it with observation semantics imported from OBI or BCO where we would define what kind of observational thresholds need to be passed. I don't think that's necessary yet, but we can make sure the comment sketches out this route as a potential extension.

The class is currently in the editors' version as:

Label: 'new particle formation event'

alternative term: 'rapid formation of liquid or solid particles from gaseous material in an atmosphere'

Def: 'An atmospheric particle formation process during which 1) gas-phase molecules form solid or liquid particles that are stable enough to continue growing in size- via the condensation of other gas-phase molecules - at a greater rate than their constituents re-enter gas phase and 2) the rate of such particle formation exceeds that which is normally present in the site of its occurrence.'

Comment: 'Note that the detection of such events, and thus the thresholds used to define them in the field, are operational and dependent on the technology and observation strategy used. The event must be intense enough, occur in a large enough area, and persist for enough time to allow the formed particle population to be observed. There are no exact or universally agreed upon thresholds for the intensity, area and duration of such events. If elevated new particle formation rate is observed in the atmosphere (away from direct sources of the related gas-phase molecules) for an operationally defined period of time, practitioners often declare that such an event has taken place.'

Subclass axioms:

'formation of particles in an atmosphere' # this is the superclass

'has part' some (
 ('formation of liquid droplets from gaseous material in an atmosphere' and 'has quality' some 'increased rate')
  or 
 ('formation of solid particles from gaseous material in an atmosphere'  and 'has quality' some 'increased rate')
)

Naturally, this class is tied to the aerosol outputs by the inclusion of the other classes as parts.

@markusstocker @PauliPaasonen if you'd like to be credited for this content (and help us trace where the knowledge came from), please add your ORCIDs or equivalent to this thread. Many thanks!

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented Apr 27, 2018

The class has a reserved (not released yet) PURL of: http://purl.obolibrary.org/obo/ENVO_01001359
It's queued for our next release, which will happen as soon as a few errors in an import are fixed. The "in editor's version' tag will be removed and this issue closed when it's out.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented Apr 30, 2018

Hi,
There are still few small problems there:

  • now in the "Def" part (1) explains "new particle formation", i.e. the very process that is referred to in the beginning of "Def", and part (2) explains the "event". This is why I would hope to put this description in two pieces, one describing the formation and the other describing the event.
  • the "alternative term" seems to describe "new particle formation", not an "event". Or perhaps this is because "rapid" stands for "enhanced". Better form for "alternative term" could be: "period of enhanced formation of liquid or solid particles from gaseous material in an atmosphere"
@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented Apr 30, 2018

Thanks @PauliPaasonen

now in the "Def" part (1) explains "new particle formation", i.e. the very process that is referred to in the beginning of "Def", and part (2) explains the "event". This is why I would hope to put this description in two pieces, one describing the formation and the other describing the event.

I'll then cast the event as a special case of the formation process. Would this work?
image

And then...
http://purl.obolibrary.org/obo/ENVO_01001365

image

And ...

http://purl.obolibrary.org/obo/ENVO_01001359
image

the "alternative term" seems to describe "new particle formation", not an "event". Or perhaps this is because "rapid" stands for "enhanced". Better form for "alternative term" could be: "period of enhanced formation of liquid or solid particles from gaseous material in an atmosphere"

The alternative term is attempting to phrase this in more BFO-compliant ways for future work / disentangling.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented Apr 30, 2018

Yes, that is correct to have the "event" as a special feature of new particle formation (NPF). However, I find the higher level terms a bit complicated to understand, since the term "particle formation" (or "formation of particles") means, at least for me, the exactly same thing, new particle formation. Additionally, new particle formation can occur also elsewhere than in the atmosphere, e.g. in a car tale pipe.
In short, the general term is typically "aerosol formation", which can be divided to primary and secondary formation. "Primary aerosol formation" means emissions of particles (solid or liquid), i.e. appearance of particles as particles from some source. "Secondary aerosol formation" is then formation of (solid or liquid) from gas phase molecules. Secondary aerosol formation can be either condensation of vapours on pre-existing particles (aerosol mass increases, number does not) or formation of new particles from vapours, i.e. "new particle formation" (total aerosol mass and number increase). Actually, I would not use terms "particle formation" or "formation of particles" at all, since they fall somewhere between "aerosol formation" and "new particle formation", and it is not clear if only "new particle formation", "secondary aerosol formation", or "aerosol formation" is meant. Sure, interpretations might be different in some communities, but I would say the above described is quite common.

pbuttigieg added a commit that referenced this issue Apr 30, 2018

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented May 3, 2018

As it is currently modeled and shown by @pbuttigieg, "new particle formation event" is a "special case" (a sub category/set/class) of "new particle formation" (and "particle formation process" more abstractly). Concretely, this means that every instance of "new particle formation event" is a "particle formation process". I am not sure this is semantically equivalent to an event being "a special feature of new particle formation".

@pbuttigieg: Does something speak against the proposal of harmonizing "particle formation" and "formation of particles" to "new particle formation"? Also, it seems that classifying "new particle formation" as "formation of particles in an atmosphere" is incorrect.

Here is a more formal version of what @PauliPaasonen suggested above:

"primary aerosol formation" 
    subClassOf "aerosol formation"
    comment "Emissions of solid or liquid particles, i.e. appearance of particles as particles from some source"
"secondary aerosol formation" 
    subClassOf "aerosol formation"
    comment "Formation of solid or liquid particulate matter from gas phase molecules"
"condensation of vapours on pre-existing particles"
    subClassOf "secondary aerosol formation"
"new particle formation"
    subClassOf "secondary aerosol formation"
    alternative "Formation of new particles from vapours"

Since "new particle formation" can occur also elsewhere than in the atmosphere, we need to specialize "new particle formation" for formation that occurs in the atmosphere and a sensible label may be "atmospheric new particle formation", thus:

"atmospheric new particle formation"
    subClassOf "new particle formation"

Now, even if we accept this structure, one remaining question is how it relates to "new particle formation event", the event. Considering what @PauliPaasonen said above:

If elevated new particle formation rate is observed [...] for a period of time, it can be called an event.

This seems to suggest that "new particle formation event" is indeed a special case of "new particle formation", namely the case in which the observed formation rate is elevated for a period of time:

"new particle formation event"
    subClassOf "atmospheric new particle formation"
    comment "New particle formation with elevated formation rate observed for a period of time"

Updated to reflect @PauliPaasonen's comments below

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented May 4, 2018

@markusstocker: This formalization of my suggestion looks very good! Two minor suggestions: 1) comment for "secondary aerosol formation" could be "Formation of solid or liquid particulate matter from gas phase molecules". This because "formation of particles" sounds like "formation of new particles", but "secondary aerosol formation" can also be "condensation of vapours on pre-existing particles", as you have in the next line. 2) "new particle formation event" should be sub-class of "atmospheric new particle formation", since the word "event" is typically applied only for atmospheric events.

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented May 4, 2018

Thanks @PauliPaasonen, I updated the comment above to reflect your suggestions.

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented May 4, 2018

Rearranging the points in #602 (comment) to respond to @PauliPaasonen

In short, the general term is typically "aerosol formation", which can be divided to primary and secondary formation. "Primary aerosol formation" means emissions of particles (solid or liquid), i.e. appearance of particles as particles from some source. "Secondary aerosol formation" is then formation of (solid or liquid) from gas phase molecules. Secondary aerosol formation can be either condensation of vapours on pre-existing particles (aerosol mass increases, number does not) or formation of new particles from vapours, i.e. "new particle formation" (total aerosol mass and number increase).

This is much clearer - I'll use this to refactor the surrounding hierarchy. We can fit in multiple processes and events in here while offering more clarity.

The only thing that's ambiguous now is that secondary formation can include simple mass-increases (i.e. no actual formation at all). Is there a source that's well known in your field I can reference for this?

Actually, I would not use terms "particle formation" or "formation of particles" at all, since they fall somewhere between "aerosol formation" and "new particle formation", and it is not clear if only "new particle formation", "secondary aerosol formation", or "aerosol formation" is meant. Sure, interpretations might be different in some communities, but I would say the above described is quite common.

I think the aerosol formation hierarchy can help us here. I'll add "new particle formation" (the process and the event) as a broad synonym of an aerosol formation event, which seems to be the most likely thing being observed.

Yes, that is correct to have the "event" as a special feature of new particle formation (NPF). However, I find the higher level terms a bit complicated to understand, since the term "particle formation" (or "formation of particles") means, at least for me, the exactly same thing, new particle formation. Additionally, new particle formation can occur also elsewhere than in the atmosphere, e.g. in a car tale pipe.

The higher level terms were defined only by their output (some particulate matter) and not by where they occur or their rates. Thus they were more general. The aerosol hierarchy you provided will likely be less confusing.

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented May 4, 2018

@markusstocker our thinking on @PauliPaasonen's comment was quite similar; I'll fold in your suggestions to the current hierarchy.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented May 5, 2018

A good reference for aerosol formation (and all other aerosol pehnomena) is "Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, 3rd Edition", by John H. Seinfeld and Spyros N. Pandis, Wiley, ISBN: 978-1-118-94740-1.

@pbuttigieg , I understand your comment "secondary formation can include simple mass-increases (i.e. no actual formation at all)", but aerosol formation is in this analogous to "ice formation" or "cloud formation" in which both it is not required that a new piece of ice or a new cloud is formed, but increase in the volume or mass of ice or cloud is enough.
And to continue with this: "aerosol formation event" is not a common nor very reasonable expression, since (secondary) aerosol formation occurs continuously in the atmosphere, because in most environments at least some condensable vapours are present. However, the circumstances for observable formation rate of new particles are encountered only every now and then, and thus the intensive formation period is typically referred to as an event.

@markusstocker

This comment has been minimized.

Copy link
Author

markusstocker commented May 7, 2018

Hi @pbuttigieg, when you are ready, can you post a screenshot as above so that we can review the current classification, before you publish the update? I think it is good if @PauliPaasonen has a chance to check it again and approve it.

Once we have the classification of the concept "new particle formation event" right, the actual fun begins, namely how to integrate the attributes and data about events we need/should/want to capture, for instance spatio-temporal locations but also the event class (according to some scheme, e.g. Class Ia, Ib, II according to dal Maso et al. or similar following another scheme) but also growth rate, etc. I think this part will need some f2f time (seems a bit challenging via GitHub and I am not sure to what extent it is strictly within the scope of ENVO). Once we have the concept for "new particle formation event" accept and published I will draft a proposal for the attributes and post it somewhere for review. Perhaps we can organize some f2f time in Hyytiälä this August.

Thanks everyone so far for all your input! Knowledge engineering and ontology is tough.

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented May 10, 2018

@markusstocker

Thanks everyone so far for all your input! Knowledge engineering and ontology is tough.

Indeed, it takes some doing; however, as we've shown here, there are quite a few subtleties out there that can confuse data integration at a later stage unless hammered out. A little trouble here can save a great deal of trouble later.

@PauliPaasonen

@pbuttigieg , I understand your comment "secondary formation can include simple mass-increases (i.e. no actual formation at all)", but aerosol formation is in this analogous to "ice formation" or "cloud formation" in which both it is not required that a new piece of ice or a new cloud is formed, but increase in the volume or mass of ice or cloud is enough.

@PauliPaasonen this is a great example of how ontologies can help clarify terminology. I'll create classes for both processes (the mass increase and the actual formation) and then a union class for the term that you use in your field. I'll add your instructive note about cloud formation and ice formation as a comment.

And to continue with this: "aerosol formation event" is not a common nor very reasonable expression, since (secondary) aerosol formation occurs continuously in the atmosphere, because in most environments at least some condensable vapours are present. However, the circumstances for observable formation rate of new particles are encountered only every now and then, and thus the intensive formation period is typically referred to as an event.

The continuous formation of aerosols would be the process. Following our discussion above, the event is indeed a period of increased rate in the process (intensive formation). I don't see an issue here, am I missing something?

@markusstocker

Hi @pbuttigieg, when you are ready, can you post a screenshot as above so that we can review the current classification, before you publish the update? I think it is good if @PauliPaasonen has a chance to check it again and approve it.

Yes. Will do.

Once we have the classification of the concept "new particle formation event" right, the actual fun begins, namely how to integrate the attributes and data about events we need/should/want to capture, for instance spatio-temporal locations but also the event class (according to some scheme, e.g. Class Ia, Ib, II according to dal Maso et al. or similar following another scheme) but also growth rate, etc. I think this part will need some f2f time (seems a bit challenging via GitHub and I am not sure to what extent it is strictly within the scope of ENVO).

Yes, this sounds like a more involved discussion best done in a teleconference. OBO resources can provide semantics for spatiotemporal entities fairly easily. ENVO can offer subclasses of the event if their are clear differentiae between them.

Once we have the concept for "new particle formation event" accept and published I will draft a proposal for the attributes and post it somewhere for review.

We can keep it on our tracker, cross-referencing this issue.

@PauliPaasonen

This comment has been minimized.

Copy link

PauliPaasonen commented May 14, 2018

The continuous formation of aerosols would be the process. Following our discussion above, the event is indeed a period of increased rate in the process (intensive formation). I don't see an issue here, am I missing something?

The problem is that for a big part of the aerosol community "aerosol formation" = "condensation growth of particles" or "primary emissions". As especially in health and legislation issues only the aerosol mass has been considered, this is logical: "new particle formation event" increases the number of particles, but the mass is not impacted before the particles have grown (due to condensation) to larger sizes. For this reason, "new particle formation" is a good term to use always when discussing the formation of new aerosol particles. Thus also the event, if meaning the period of intensive formation of new particles from vapours, should be referred to as "new particle formation event", not "aerosol formation event" or "formation of aerosols".
I guess this looks a bit ridiculous getting stuck with such detail, but "aerosol formation" is a tricky word, meaning a bit different things in different communities.

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented May 15, 2018

@PauliPaasonen

I guess this looks a bit ridiculous getting stuck with such detail, but "aerosol formation" is a tricky word, meaning a bit different things in different communities.

It's a little painful, but this is the right place to figure out a way to expose this ambiguity coherently. Thank you for your engagement. We're unlikely to get this in a perfect state, but we can get close to communicating what's going on in the heads of scientists working with these phenomena.

pbuttigieg added a commit that referenced this issue Aug 21, 2018

@pbuttigieg

This comment has been minimized.

Copy link
Member

pbuttigieg commented Aug 23, 2018

Updates pushed to editors' version, release in a few days.

@pbuttigieg pbuttigieg closed this Mar 18, 2019

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.