@kurtwalker I also had this question and agree it is not obvious. The fact that the observation vs forecast comparison is at different refresh frequencies (based on how frequently the model prediction refreshes) makes this a less useful feature. Of course the models that refresh more frequently will perform better. The function only tests short term prediction accuracy in that case. I would prefer a standard time out comparison based on the slowest refresh rate. Or better yet to be able to control the prediction time somehow! Maybe a button for 1 day or 3 day. That could be a really wonderful feature for comparing real world model performance.

@idefix37 It is also unclear how the median error thumbs are calculated (far right on the obs vs forecast tab, thumb indicating the lowest mean error). Is that median error for the past hour? week? all recorded time at this station/model combination?

Thanks for this wonderful app. I use it everyday!

@Bartcola, the scale is for PM 2.5 not the air quality index (AQI, 0-500 scale). The AQI index is calculated with other pollutants as well as PM 2.5. The scale for PM 2.5 is in micrograms per meter cubed and is one factor used to calculate the AQI. For more info check out this site https://www.airnow.gov/aqi/aqi-basics/

Thanks for this wonderful App! I am a big fan.

In short, I'm hoping someone would develop a plugin or feature that would return a text forecast to a sat messenger when requested. In areas without cell service this could be a fantastic service. For example, if I sent a message with my lat/long, model requested, and requested information (e.g. surface wind) the service would provide a text output of the requested point forecast. Most satellite messengers have weather services, but they lack the resolution I need for an accurate forecast in my area. I'm particularly interested in the HRRR model for Alaska. Hoping someone with more programing experience will make this happen so I don't have to :). Thanks!

Love the distance and planning tool, and the app in general! Great work! It would be useful to be able to change the forecast model from ECMWF. My use case is traveling in a kayak, but a it would be useful for any small vessel to have models with higher spatial resolution.
Thanks for all your work on this app!

@kurtwalker I also had this question and agree it is not obvious. The fact that the observation vs forecast comparison is at different refresh frequencies (based on how frequently the model prediction refreshes) makes this a less useful feature. Of course the models that refresh more frequently will perform better. The function only tests short term prediction accuracy in that case. I would prefer a standard time out comparison based on the slowest refresh rate. Or better yet to be able to control the prediction time somehow! Maybe a button for 1 day or 3 day. That could be a really wonderful feature for comparing real world model performance.

@idefix37 It is also unclear how the median error thumbs are calculated (far right on the obs vs forecast tab, thumb indicating the lowest mean error). Is that median error for the past hour? week? all recorded time at this station/model combination?

Thanks for this wonderful app. I use it everyday!

Thanks for this wonderful App! I am a big fan.

In short, I'm hoping someone would develop a plugin or feature that would return a text forecast to a sat messenger when requested. In areas without cell service this could be a fantastic service. For example, if I sent a message with my lat/long, model requested, and requested information (e.g. surface wind) the service would provide a text output of the requested point forecast. Most satellite messengers have weather services, but they lack the resolution I need for an accurate forecast in my area. I'm particularly interested in the HRRR model for Alaska. Hoping someone with more programing experience will make this happen so I don't have to :). Thanks!

Love the site/app I am constantly impressed with the information provided. In the way putting my vote in, I'd love to see a tide chart like the one user lumpet one and user wharfRat posted about. Also as others have mentioned the meter to foot conversion is not correct for tidal height, e.g. 0.23 meters is not 0 ft. Another thing that would be lovely is a graphical interface like this:

deepzoom.com

where tidal current stations have a graphical representation of direction of flow and water speed. Please and thank you! Keep up the good work!

@Bartcola, the scale is for PM 2.5 not the air quality index (AQI, 0-500 scale). The AQI index is calculated with other pollutants as well as PM 2.5. The scale for PM 2.5 is in micrograms per meter cubed and is one factor used to calculate the AQI. For more info check out this site https://www.airnow.gov/aqi/aqi-basics/