2021 data release date? and 2020 annual data

[Anu Kramer]

Hello lovely CFO community!

I see in the documentation that the annual data are based on Sept/Oct, but when I download the data from the website interface, it is the for the summer of 2020. I'm not able to access the computer that I normally use to download via the API to check what is available there for 2020, so my apologies if additional data are available there, but could someone please let me know:

• Is the annual data for 2020 available for download via the api?
  • if so, are those data from sept/oct of 2020?
• When will the 2021 data be released?

Note that I'm specifically interested in canopy height and canopy cover.

Thanks so much!

     - Anu

[Anu Kramer]

To follow up, I was able to get onto the API, and it seems that there aren't any annual layers; just Summer and a smattering of Fall and Spring, which are insufficient for my needs.

It looks like I'll need 2021 data for canopy height and canopy cover, and while there are some layers currently on the API for Central Coast, Piliga, and South Mountains, these are not loading in ArcGIS. Does anyone know when the rest of 2021 will be released or whether there's a fix to get these layers to display (I'm getting the error "The parameter is incorrect")?

Thanks,

         - Anu

[Anu Kramer]

Another followup - I got these 2021 layers to load in Google Earth, and it looks like they're all in Australia, which is a new geographic range from previous years!

Any info on when 2021 will be available for California?

Thanks,

         - Anu

[Christopher Anderson]

Hi Anu!

I'll try and clear up some confusion that I suspect is driven by how we process some of our datasets.

Annual datasets

The 10m layers are updated on an annual basis and labeled as "Summer." This is a bit imprecise, because they are a composite of multiple satellite observations and an ensemble of model predictions across time periods that span from the Spring to the Fall. The image compositing is performed to create cloud-free and gap-free mosaics, the model compositing to reduce bias and reduce the effects of seasonal observation variance (e.g. accounting for varying phenological and observation conditions). We also filter the satellite observations to ensure any early-season fires are excluded. Together, the output maps represent the pre-fire season conditions of each Forest Observatory metric, which we represent as annual data products.

Our partnership with Planet provides access to high resolution imagery for 2016 and 2020, though the satellite measurements were made during shorter periods of time. The 2020 data were collected in Spring, as we had to acquire and process the data prior to the CFO launch in Fall 2020. The 2016 data were collected in Fall a) to ensure gap- and cloud-free cover (as fewer measurements were available before then) and b) to match the most recent available year of LANDFIRE data at the time.

2021 data release

This release is a long time coming, and I apologize for the delay. It's the result of many factors, and I'll spare you most of the details. We hope to have the 2021 layers released by September or so, which will include a refresh of each forest structure layer in addition to a series of new layers that we'll announce when they're ready.

I could use your input on something! One of the delays is due to us refactoring a lot of our data processing and modeling pipelines, and the output models are now higher precision than the 2016-2020 releases (as measured by increases in r-squared). We're debating between the value of continuity vs. the value of accuracy. Is it more important for you to have 2021 data that is most consistent with the 2020 data, or 2021 data that is as accurate a measurement we can make?

Australia data expansion

I see you've sleuthed your way into discovering the data from New South Wales. This was another cause for our delay. We haven't made any announcements, but this work was done to the Minderoo Foundation's Fire and Flood Resilience initiative, and you can see some news coverage of the pilot via Channel 9 News (note: we weren't paid all that money; that's the amount for the whole initiative). We were quite excited to have the opportunity to prove that the technology approach scales, and can be modified to adapt to new modeling contexts (we mapped a series of fuel models that fit into several Australian fire modeling contexts in addition to supporting the USFS fuel modeling approach).

We'll be in touch as we come up with more concrete plans for these releases. In the meantime, thanks for your patience, and thanks for being an active Forest Observatory user!

[Anu Kramer]

Greetings,

Thanks so much for your detailed response! I just realized that I didn't reply to your question, for which I apologize! I imagine that you've already called the shots on this one, but in case you're still looking for input: I'm involved in a bunch of projects, and backward consistency is important for some, while others would benefit from an accuracy boost. I guess ideally I'd be able to access both, but I realize that's super confusing.

A few questions for you:

- I'm getting an error when I try to fun forest.search:

forest.search(category="Vegetation", year=2016, timeOfYear="Summer")
2022-10-13 11:02:43,086 WARNING cfo.utils [search] | b'{}\n'
b'{}\n'
- I can't check the api due to the above error, but I'm not seeing the 2021 data on the online map platform, and I don't see any notes about the release in this group. Do you have any more updates on when it might be available?

Thanks so much,

       - Anu

[Kamini Yadav]

Dear Christopher,

I am also interested to know about the release of 2021 data.

Kamini

--
To unsubscribe from this group and stop receiving emails from it, send an email to community+...@forestobservatory.com.

--

=====================================

Kamini Yadav, PhD ORISE Post Doctoral Fellow US Forest Service Pacific Southwest Research Station p: 575-621-8057 kamini...@usda.gov 4955 Canyon Crest Drive Riverside, CA 92507 www.fs.fed.us Caring for the land and serving people

[Durrell Kapan]

Hi all,

Count me in as also interested in 2021 data.  For our purposes backward compatibility is critical (as we compare between years).  

Any update on the release of data and whether or not it will be 'backward compatible?'

Thanks, CFO, for your excellent tools!

All the best,

Durrell 

Durrell D. Kapan, Ph.D. 
Senior Research Fellow 

Institute for Biodiversity Science and Sustainability

California Academy of Sciences

www.calacademy.org

[Anu Kramer]

Hello CFO? 

If I don't hear anything from you soon we may have to pass on using the data, which would be an enormous bummer!

If anyone on this forum knows a better way of contacting CFO, please let me know.

Thanks!

       - Anu