2) Be cautious in extrapolating from one location to another.  Temperature, and especially rain, estimates for the nearest Radar site may not accurately reflect conditions at your location.
 

3) Each table or chart shows the time when forecast values begin.  This time should be 2AM or 2PM of today's date.  If an older date is showing, try selecting "Refresh" from the Internet Explorer View menu (hitting the F5 function key does the same thing)   This forces the browser to retrieve the file from the web instead of using a older version of the file stored on your computer from a previous browsing session.
 
    Not all pages are updated every day.  Only those Radar web pages that will be affected by new weather information are updated each day.  For example, the Flyspeck respray interval tables for August will not change until August dates come into the 10 day forecast range.  So those tables do not begin updating until late July.  For the same reason, once a table or chart has all of the data it will ever need, it stops being updated.

   Orchard Radar is updated three times a day, 7 days a week, starting at approximately 4:20am, 5:30am, and 3:50pm.  The update times for each site are shown below the "32 Day Rain Chart".
 

4) Charts and tables are designed for a minimum screen resolution of 1024 x 768.  Be nice to your eyes, read optimizing your Orchard Radar display.
 

5) To better understand how the estimates are made, read the background information pages. Each model output in table format has a link to the background page for that model.

Models as Orchard Decision Aids
 - presentation at 2001 New England Fruit Meetings by Glen Koehler

   'Orchard Radar' is a catchy name for a simple concept: using desktop computers and the internet to acquire updated weather observation and forecast data, feed it into apple pest management and horticulture models, and then use the internet to distribute the output to growers.

     I use the term "Radar" because it implies an appropriate way to view these products as tools that help you gauge the relative nearness (in time) or size (severity) of something.   Orchard Radar gives you early warning about apple pest risks similar to the way true radar indicates the nearness and size of an approaching object.  These products are only like radar in this conceptual sense.  There is no direct use of radar technology in translating the weather data into orchard estimates.

   Growers are familiar with relationships between weather variables such as temperature and rainfall and events in the field.  Some of these relationships are simple, like knowing that warm weather leads to earlier bloom or maturity date.  Others are more detailed, such as the interactions of temperature, rain, and leaf wetness on apple scab infection potential. 

    Many such relationships have been formally quantified and published in scientific literature.  Others are used as informal "rules of thumb", such as "Imidan coverage on foliage loses effect if there is two inches of accumulated rain during the 10 days after application, or if that does not happen, after 1.5 inches accumulated rain up to a maximum of 14 days."  The models that make up Orchard Radar consist of both formal and informal relationships.

   Those relationships have been known for years, but the difficulty in collecting and processing weather data on individual grower computers has prevented full use of the available information by orchard decision makers.   The emergence of the internet into mainstream use has provided a path by which a single computer can handle the large volume of weather data for multiple sites, process that data, and then quickly, inexpensively, and automatically broadcast the estimates to growers.

     As with any decision-making aid, common sense is required to use Orchard Radar tool prudently and effectively.  There is no guarantee for the validity of each model for each location and situation.

     These are experimental tools still undergoing development.  Grower feedback is important and appreciated.  Are there better ways to present the information?  Which Orchard Radar products are useful, which are not?

     The intention in providing Orchard Radar is NOT to have a machine attempting to make farm management decisions. The intention is to improve the quality of the information base upon which YOU make decisions.   In no way is Orchard Radar intended to reduce reliance on your existing means of arriving at decisions.  As with any other source of information, you should look at the Radar estimates as one more piece of the puzzle.  

   My hope is that with experience in referring to the Radar products and how they relate to the real world in your orchard, you will find that they are a useful addition to your Management toolbox.  Like any tool (including pesticides), the Orchard Radar products are only useful to the degree that they are used properly.

     The weather data are tailored for specific orchard locations by interpolating between values received from surrounding National Weather Service reporting sites.  Be cautious in extrapolating from another site to yours. Some rains are caused by the movement of large fronts that deliver consistent amounts of rain across large areas.  Other rains, such as summer thunderstorms, can be a very localized event, causing a nearby site to have estimates that do not match conditions at your location.     

    In addition to interpreting weather observations, the Orchard Radar also uses forecast data.  Weather forecasting is less than perfect, but it is much better than using simple climatic averages.  Rain forecasting is significantly better than climatology out to about 5 days. Temperature forecasting is significantly better than climatology out to about 7 days. 

   Orchard Radar uses forecast information out to 10 days, but the last 3 days are essentially just a trend back to the climatic average, not daily specific forecasts.  Beyond 10 days it uses climatology to provide useful guidance for ‘about when’ things are likely to happen. 

   As the date of a predicted event approaches, climatic and forecast values used in the estimate are replaced by observed values.  Thus, the nearer the event, the better the estimate.