Add support for phenology observations in Austria

[amotl]

Dear @HannesOberreiter,

at ¹, we are presenting a new subsystem for the phenodata program to export data into SQLite databases for easy consumption in downstream applications. I think it could be sweet to follow up on this by looking at whether the Austrian government also provides unrestricted access to phenology observation data.

In Austria, there are both the Phenowatch, and the Pan European Phenological database (PEP725) projects being managed by ZAMG / GeoSphere Austria. While they aimed to make the data accessible in an unrestricted manner,

• https://www.phenowatch.at/aktuell/phaenologie-spiegel
• https://www.zamg.ac.at/cms/de/klima/klima-aktuell/phaenospiegel/jahr
• https://www.phenowatch.at/qualitaetssicherung-von-citizen-science

Wo werden Daten und Metadaten öffentlich zugänglich gemacht?

Die Daten werden auf den Webseiten grafisch aufbereitet dargestellt und sind außerdem nach erfolgter Qualitätskontrolle auch über die Paneuropäische Phänologische Datenbank (www.pep725.eu) für Forschungszwecke frei verfügbar. Darüber hinaus kann jeder User der App die Daten seiner gemeldeten Beobachtungen direkt in der App oder auf der Webseite herunterladen.

Wo werden die Ergebnisse öffentlich zugänglich gemacht, sodass sie auffindbar, weiterverwendbar, nachvollziehbar und transparent sind?

Ergebnisse werden soweit möglich auf den Webseiten des Paneuropäischen Netzwerkes (www.pep725.eu) und der Webseite des Phänologischen Netzwerkes (www.phenowatch.at) und der Webseite der ZAMG (https://www.zamg.ac.at/cms/de/klima/klima-aktuell/phaenospiegel/jahr) veröffentlicht.

presumably under the umbrella of PEP725,

The main objective of PEP725 is to promote and facilitate phenological research by delivering a pan European phenological database with an open, unrestricted data access for science, research and education (datapolicy).

I think it may be that they could not hold up to it.

NOTICE: Currently, this website contains only observations until 2016 - if you need more up-to-date records for your research please get in touch with us. The available records/country are always shown on the statistics page.

-- http://www.pep725.eu/data.php

Maybe you can conduct some research, and share back corresponding pointers with me? It is easily possible that I missed to find the corresponding download space or API interface. Thank you so much!

With kind regards,
Andreas.

Footnotes

1. https://community.hiveeyes.org/t/phenodata-ein-toolkit-zur-beschaffung-und-verarbeitung-von-open-access-phanologiedaten/2892/57 ↩

[HannesOberreiter]

Hi,

recently I did a similar research for some open pheno data, to integrate into my app. The PEP resource was also the best
I could find. I would expect that the phenowatch data is included in the pep dataset. Anyway their data policy is rather restricted, at least for my semi commercial purpose, thats why I never followed up on it and never registered for their service, which you need to do before getting the latest data.

Maybe offtopic, but currently my idea is to use iNaturalist API (https://api.inaturalist.org/v1/docs/). In the portal users can add to plant observations the current phenology state and you can filter in the API for this state.

Sadly not many users use this feature, so my idea was to first write a pull request for the iNaturalist app. That users can also set the phenological state in the mobile apps (which is not possible currently). Which would probably increase the ratio users would select a state.

Cheers
Hannes

FYI: (No pheno data) For Austria in specific, if try to find open source resources on following homepage which was (sometimes) helpful: https://www.data.gv.at/

FWIW: A few years back I had high hopes for an intereg funded project but there was nothing emerging from it (https://keep.eu/project-ext/1307/).

[amotl]

Dear Hannes,

thanks for your swift response. iNaturalist has already been mentioned on our forum ¹, and I took a short glimpse into it. I think I remember it as the go-to archive about what it advertises itself with:

The iNaturalist Open Dataset is one of the world’s largest public datasets of photos of living organisms, containing over 70 million photos ².

On the other hand, I have not been aware that they or their communities are also running continuous or recurring observations like the DWD and their observers are conducting/managing - I must have missed this detail.

Anyway their data policy is rather restricted [...], thats why I never followed up on it and never registered for their service, which you need to do before getting the latest data.

Yeah, it's sad. Access should be unrestricted.

So, we don't have any other promising data source candidates for Austria, right? Maybe we should reach out to ZAMG/Geosphere and humbly ask about the current state of the data management, and whether there would be room to improve?

On the iNaturalist API, do you think it would both fit into and bring value to the phenodata program I am conceiving here? I took a short glimpse, and found entities like species, phases, places, observers, and observations, which feels familiar. On top, there is even a ready-made client library called »pyinaturalist« ³, which looks pretty comprehensive, and would definitively save some keystrokes.

With kind regards,
Andreas.

Footnotes

1. https://community.hiveeyes.org/t/neue-bee-counter-software-auf-basis-von-computer-vision-und-machine-learning/868/9 ↩

2. https://github.com/inaturalist/inaturalist-open-data ↩

3. https://pyinaturalist.readthedocs.io/ ↩

[HannesOberreiter]

Hi,

On the iNaturalist API, do you think it would both fit into and bring value to the phenodata program I am conceiving here?

I would think so, although the interpretation must be seen different. As it is not "first" flowering observed - as typically for phenodata - it a current state of the flower, therefore flowering can happen over multiple months for certain taxa. The API itself is straightforward (but often not well documented and buggy, but hey at least they offer it).

Example request: Get me all Corylus avellana which are flowering in the year 2023.

https://api.inaturalist.org/v1/observations?taxon_name=Corylus%20avellana&year=2023&term_id=12&term_value_id=13&order=desc&order_by=created_at

I don't think you need an extra client layer in between for this task.

So, we don't have any other promising data source candidates for Austria, right? Maybe we should reach out to ZAMG/Geosphere and humbly ask about the current state of the data management, and whether there would be room to improve?

Probably, on that note I think the Geosphere change was rather recently. At least for me it is new. PEP I could see would be difficult, as there are multiple parties involved with the data, but phenowatch.at could maybe open up.

FYI: iNat are also working on an V2 API, which in addition lets you define the fields returned to save some bandwidth. Which I currently use for following upcoming event: https://hannesoberreiter.github.io/inat-austria-city-challenge-2023/

Cheers
Hannes