National Digital Forecast Database (NDFD): Is it the right forecast to use?

[czenzel]

Hi,

I have been using the National Digital Forecast Database (NDFD) from the National Weather Service for a few months now on my web site to get weather for my local area with my weather station. I have found it accurate, but sometimes it gives too many specifics and calculates precipitation almost everyday when their isn't even a millimeter of rain.

Does anybody recommend the use of or against the use of the NDFD? Is there any other forecasting models I should be considering for pulling data from for local analysis?

I know of the AVNMOS, GFS models, but they do not provide the time periods that the NDFD does and they are really specific on the weather station. NDFD provides a gridded-base to allow latitude and longitude input.

Thanks,
Christopher Zenzel

 

[rdale]

The NDFD is not a model, it's the official forecast from the NWS. A meteorologist (or several) looked over data and came up with that as their best outlook for conditions, so they incorporated all of the available models and their knowledge to come up with it.

 

[czenzel]

The NDFD is not a model, it's the official forecast from the NWS. A meteorologist (or several) looked over data and came up with that as their best outlook for conditions, so they incorporated all of the available models and their knowledge to come up with it.

So, NDFD is just a gridded forecast based on models and forecasters by NOAA meteorologists. Why do text forecasts still exist also (for each county/zone)?

Do you think NOAA will ever upgrade to just use NDFD, or will we always have multiple forecast types (text, gridded, etc)?

Thanks,
Christopher Zenzel

 

[rdale]

Text will likely always have to exist for ease of use... Many media outlets / EMAs / public safety offices / etc still get the text feed, so they'll want to see that zone forecast as a quick look.

I suppose at some point they might drop the text, but when that happens then the private sector developers will likely come up with a program/web interface to convert NDFD to text.

The last thing I want a radio DJ to do is click all over the place to get the forecast...

 

[Tyler Allison]

I suppose at some point they might drop the text, but when that happens then the private sector developers will likely come up with a program/web interface to convert NDFD to text.

yes they will. Create a void...fill a void.

 

[rdale]

God Bless polygons and the void they created!

 

[Zack Hargrove]

Text forecasts will always be very important because it is imperative that we know why a forecaster is forecasting certain values. When viewing forecasts in areas we do not live in or areas we are not familiar with, the text forecast gives us insights on features and terrain that a mere group of models or graphical forecasts do not show (the mountains are a good example of this... I live in Asheville, NC and the models are constantly wrong in the winter because of mountain issues with up slope snow, etc...). We should always look at models and make our own decisions, but checking the text forecast should always be number one priority rather than merely looking at graphical forecasts. The amount of information and the quality of the information is without question more valuable. The why can be just as more important if not more important than the what.

 

[rdale]

Text forecasts will always be very important because it is imperative that we know why a forecaster is forecasting certain values.

I'm not sure I follow what you are trying to say... The text forecast comes directly from the NDFD forecast. There is no forecaster "why" in the forecast, just a text format of the NDFD. That's why you get some strange things like:

Chance of rain, sleet and snow until 2am.
Sleet and snow with a chance of rain until 4am.
Rain and sleet and snow until 6am.

etc. The humans can look over them and tweak if needed, but in the big picture there is no "why" -- just "here is NDFD in a text format."

 

[Jeff Duda]

Text forecasts will always be very important because it is imperative that we know why a forecaster is forecasting certain values. When viewing forecasts in areas we do not live in or areas we are not familiar with, the text forecast gives us insights on features and terrain that a mere group of models or graphical forecasts do not show (the mountains are a good example of this... I live in Asheville, NC and the models are constantly wrong in the winter because of mountain issues with up slope snow, etc...). We should always look at models and make our own decisions, but checking the text forecast should always be number one priority rather than merely looking at graphical forecasts. The amount of information and the quality of the information is without question more valuable. The why can be just as more important if not more important than the what.

I think you might be referring to the Area Forecast Discussion product. That's where you get the "why" of the forecast. But the AFD is not a text forecast product of its own. It's just a text file that explains the forecaster's reasoning.

 

[Zack Hargrove]

I apologize... I see that now... I had misunderstood the original post. That's what I get for not reading the whole post.