Weather Data For Observations

I was wondering if you’s guy’s can implement a feature to include weather data to be displayed with observations? Like temperature, humidity, wind speed & direction, overcast, sunny, rainy, foggy etc and moonphase?

I think it would be helpful for gathering data and research. Anybody else think this would be an awesome new feature?

  • Yup
  • Nope

0 voters

All of those variables (and more) can be added already via Observation Fields. They aren’t necessarily standardized across all taxa, and many are redundant and/or project-specific. Even if the fields were standardized, many people won’t have the means nor interest to collect this data while snapping a photograph, especially casual users or even professional biologists (I generally don’t carry a thermometer, hygrometer and anemometer with me). The data would be virtually meaningless if say, half of users left fields blank, most of the rest guessed the temperature, or used subjective values (my light rain may be your heavy mist), and only 5% took quantitative measurements.

5 Likes

Yes, temperature of great interest here the last few months, as it is extremely high on a certain ridge, I believe up to ten degrees celsius higher than reported by local met. Service for this area…as they don’t have a weather station in this locality
EDIT and I recently realised there is a temp sensor in my camera. And the elevation, which is relevant to temp as well

2 Likes

Fields can capture this, and you can create fields if the ones you want don’t exist. You can also (and should) add observation information like this to the description.

Retrospectively, for places that have a lot of weather recording stations etc, you can use the pin location and date/time to look up recorded meteorological readings. It won’t of course capture micro-climate data, but for that you would need suitable equipment anyway.

2 Likes

Weather data is extremely unreliable, even after the fact. In most cases it’s interpolated from the nearest weather station, and even over close distances it can be completely different. You also have to ensure that whatever service is using the actual nearest station, not just the most accessible.

At the moment I’m working in an island in northern Vietnam. The nearest actual weather station is the Hai Phong airport, but I can select Hai Phong or Ha Long City as my location on the better services. The two weather reports are very different, and if you dig into the data you find that the Ha Long City weather is interpolated from Hai Phong Airport data, and even the Hai Phong City and Hai Phong Airport data is slightly different as one is actual data and the other is modified data. Both of them are completely wrong for where I am, just a few tens of kilometers away.

Where I physically am can have the difference between thundering rainstorms and heavy wind that drive everything into hiding and bright, calm sun in just a kilometer or two. Indeed, depending on the time of year I can predict certain weather changes to within a couple hundred meters when I’m driving back and forth.

To make matters worse, most of the weather data apps and websites use for my area actually come from Hanoi which is used as the “Northern Vietnam” base for predictions, despite being in a wildly different type of environment than much of the rest of northern Vietnam and has weather totally unlike everywhere else.

This is a common type of problem and is even worse in may areas.

I’d absolutely fight against any automated addition of such data as it would be nearly impossible to have it accurate for the specific location of your observation.

However, if you want to add what you actually see and experience during the observation that would be excellent additional data and far more accurate than an automated system. That said, a certain level of consistency and rigor would need to be applied: is if foggy, or low overcast? What about mist vs fog? Where is the line between rain and drizzle, etc. Those sorts of things can be important distinctions, especially when it comes to insects.

8 Likes

Yeah, I lived in Palmerston North for a while, a small city on an alluvial plain at the base of a mountain range and a gorge running through it. Weather would vary tremendously, with as much as 20km/h difference in wind speeds alone from east to west.

Fwiw I don’t think OP is suggesting auto application of weather data, but user perceived data. It’s really starting to stray too far from the iNat mission though.

Only reason I brought up getting data from recorded weather stations is to show that anyone wanting to retrospectively get an approximation of conditions can do so via that route.

3 Likes

I forget about the observation fields. Thank you!

1 Like

The sites I am interested in the temp of are up to 10degrees C higher than any records I can access form NIWA or metservice, which is why they are of interest. I dont have time to find the temp data for each file, and dont always know which ones will be of interest, which is why it would be great if iNat would read and display the data for me:)

The temperatures are automatically captured by my Olympus TG5 camera, I recently discovered.

But if your photo has been tagged with the met data then anyone interested in them CAN get them from the iNat image info page… You don’t need to know which ones they are interested in. Anyone wanting that data in a serious way can script or programmatically extract it, and theoretically could even extract and add as fields

My desire would be for me to be able to see that info automatically, in the same way as location and time. I am not primarily concerned with other people’s uses.

This is why I thought you meant other people being interested in your met data…

2 Likes

Oh yes, I see…well others too of course:)

1 Like