Background for
"Weather Influence on Postbloom Apple Thinning" table and chart.
 

    This page explains the methods used to interpret the influence of weather on sensitivity of apple fruitlets to thinning agents applied between petal fall and 18mm diameter.

Sources for model logic:

Pink to Petal Fall predisposing weather factors:

    Apple Thinning Guide by Philip G. Schwallier, Clarksville Hort. Exp. Sta., Michigan State University Cooperative Extension, 1996. 
 

Weather influence on sensitivity to postbloom thinning agents:

     2000-2001 New England Apple Pest Management Guide, University of Maine Cooperative Extension.  Section on "Chemical Thinning" edited by W.R.Autio. 

     2002-2003 Pennsylvania Tree Fruit Production Guide, Section on "Chemical Thinning of Apples" coordinated by R.M.Crassweller.  Online at  http://tfpg.cas.psu.edu/part1/part14c.htm 

     Environmental Factors Affecting Fruit Set and Chemical Thinning of Apples,  by R.E. Byers, Virginia Polytechnic Institute and State University, 2000.

     Light Regimes within standard size apple trees as determined spectrophotometrically.  Proc. Amer. Soc. Hort. Sci. 93:1-6, N.E. Looney, 1968.

    Pest Management Guidelines for Commercial Tree-Fruit Production 2002, Cornell University Cooperative Extension.  Section on "Growth Regulator Use in Apples", authored by T.L. Robinson, J.R. Schupp, and S.A. Hoying.

   Review comments by Dr. Duane Greene, University of Massachusetts, May 2008.

Procedures:

        Weather data values (hourly night temperatures, hourly solar radiation and % of full potential radiation) used for this model are from SkyBit Inc.  However, the model can run with temperature and sunlight data from any source.   The SkyBit data provide observed and forecast data in the same file.

        The model is updated twice daily with updated observation and forecast data.  Heat units are the average number of degree hours above 50F for the 12 hours between 7pm and 7am each day.  Night temperatures above a 50F threshold are associated with higher sensitivity to thinning agents.  Values displayed in table are rounded to whole numbers, but true values are used in calculations.  

        Shading (decreased sunlight) is also associated with higher sensitivity to thinning agents.  The % full sunlight data are stated as the inverse value, i.e. % shading so that higher values for both factors (temperature and light) are associated with higher sensitivity to thinning agents.  This makes it easier to follow the model logic, read the table and assess how each day's weather influences sensitivity.

--------------------------------
Rankings for Night temperature and Shading %:

0  -- (night temperature only) Very strong temperature influence
        for decreased sensitivity
1  -- Stronger influence for Decreased sensitivity
2  -- influence for Decreased sensitivity
3  -- Intermediate, no particular influence 
       towards either decreased or increased sensitivity
4  -- influence for Increased sensitivity
5  -- Stronger influence for Increased sensitivity
- - - - - - - - - - - - - - - - - 
6 -- (night temperature only) risk of Excessive Thinning
       if combined with intense cloudiness
7 -- (night temperature only) risk of Excessive Thinning
       unless combined with above average sunlight %

         For each 24 hour period, the night time degree day base 50F
accumulation is calculated and ranked by the scale below:

Average number of
degree hours base 50F
from 7pm -- 7am           Ranking

0			0
1			0.5
2			1
3			1.5
4			2
5			2.5
6.5			3
8			3.5
10			4
14			5
19			6
25			7

        To avoid jumps in ratings between similar values on different sides of a category

boundary, rankings are assigned along a continuous gradient.

         A simple formula assigns rankings for nighttime DD50 accumulations
that are <= 4.  The formula is:  Rank = Night DD50 / 2.

        A linear regression is used to assign a ranking along a continuous gradient 
to nighttime DD50 accumulations between 4 and 25.  

----------------------------------------

For each 24 hour period, the % full sunlight night value is calculated and ranked by the scale below:
 

% shading 

(= 100% - % full sunlight)		Ranking
<= 10				1
15				1.5
20				2
25				2.5
30				3
35				3.5
40				4
45				4.5
>= 50				5

	
     To avoid jumps in ratings between similar values on different sides
of a category boundary, a linear regression is used to assign a ranking
along a continuous gradient  to each day's shading % if the value is
between 10% and 50%.  

    Values below 10% or above 50% are assigned a value of 1 or 5,
respectively.

     Overall, the % shading value generally averages around 30%, 
but few individual days are at the average value.  The average derives
from a majority of sunny - partly sunny days at 80%+ full sunlight,
(= 20% or less shading) combined with less frequent cloudy and sometimes
rainy days with 40%+ shading.  
 
A very sunny, blue-sky day is 90%+ full sunlight (<=10% shading).  

    Overcast days usually have 50% or less full sunlight.  A heavily 
overcast-all-day rainy day has <40% full sunlight (> 60% shading).

Summarization and statement selection:

a) Rankings of night temperatures for the first three nights
following a morning thinner application on the stated date
are averaged.  

b) Rankings of shading % for the application day, 

and the two following days are averaged (average of 3 values).

c) The average shading and night temperatures are not given 
equal weight in the final assessment.  Night temperatures are given double weight.  
Thus, the average is of three values:  
     1. average of 3 shading ratings
     2. average of 3 night temperature ratings
     3. average of 3 night temperature ratings counted again.

   Temperature is given twice as much weight for two reasons.  

       First, with cool temperatures, thinning response is poor regardless of light levels.
With very high temperatures, thinning response can be excessive regardless of 
light levels.  Thus, temperature is the more important factor.  

       Secondly, a chart of the influence of temperature and light intensity on thinning 
response in the Cornell Pest Management Guidelines for Commercial Tree-Fruit
Production shows temperature influencing thinning response across four categories:
poor, moderate, good, excessive. 
 
        The light level influence in that chart is across two categories within each 
temperature rating.  Thus, the range of the light influence is half that of the
temperature influence on the final combined rating of thinning response.

        The influence of a single day's rating for each factor (sunlight, night temperature)
on the final composite thinning sensitivity rating is as follows:

        
Night temperature:  0.33 * 0.67 = 22%

        (each value is one of three that goes into the average for that factor, 
         and that average is given 67% weight in the combined average)

        Shading:  0.33 * 0.33 = 11%

        (each value is one of three that goes into the average for that factor, 
         and that average is given 33% weight in the combined average)

         The average of a and b is evaluated by the following scale:
If the average is <1.75, "Stronger influence for Decreased sensitivity",
If the average is > 1.75 and <=2.5, "Decreased sensitivity",
If the average is >2.5 and < 3.5, "Intermediate",
If the average is >= 3.5 and <=4.25, "Increased sensitivity",
If the average is >4.25 and < 5, "Stronger influence for Increased sensitivity"
If the average is > 5,  "EXCESSIVE THINNING RISK"

        The intent is to highlight those conditions most likely to lead to
increased or decreased sensitivity to thinning agents.

        Because trees can be predisposed to high or low thinning response
due to factors beyond the scope of this model,
the model statements do not state expected thinning outcomes 
(such as 'poor', 'moderate', or 'good' thinning) in the summary statements.  

        Instead, the statements are stated in terms of the sensitivity
to thinning agents expected relative to other weather conditions.

Range of dates used to calculate summary statement
 for each potential application date:

        While a longer period of days before or after the application day
 could be used in calculations, that would present two problems.

1) The more values used to calculate an average, the more likely that 
the average tends towards a central value.  

        Using more than 3 days of shading, or 3 days of degree nights,
 makes it less likely for the resulting  "averages of the averages" to deviate
 from the center value.  Yet grower and researcher experience shows that 
weather conditions often do influence the efficacy of thinning agents.

        The model could be adjusted for this by broadening the criteria used to 
assign summary statements about thinning sensitivity.  However, doing that is

intuitively difficult as it would require much broader ranges for assigning summary
statements than the category assignments give to values for individual days. 

2)  Even if broader ranges for the average values were identified, 
doing so assumes that conditions immediately before or after have no more

influence than conditions that occur longer before or after the thinner application.

That does not seem like a valid assumption.  While a time weighting scheme 
could be devised, these effects are only beginning to be understood at a
crude level, and there is no available basis for building such complexity
into the model.

       Therefore, because of the problems incurred by averaging too many
days into the summary assessment,  just the days with the presumed
strongest influence on a thinner for a given application date are included
in the ratings, i.e. night temperatures and daytime shading on the day of thinner

application and the next two days.  

        In a previous version of this thinner sensitivity model, shading for the two days

prior to application was averaged along with shading on the day of application, and on

the following two days.  This was based on the idea that shading on the two days prior

to thinner application was thought to reduce cuticle thickness and ability of the foliage

and fruit to absorb a chemical thinning agent.  

        In addition, shading on the two days prior to thinner application were thought

to have an effect on the accumulation of carbohydrate reserves that affect the capacity

of apple trees to overcome  the physiological stress of the thinner.

       Consideration was also given to the idea that the model should only represent 
the post-application impact of shading on net carbohydrate storage as the most immediate 
impact, and to leave out consideration of sunlight % in the two days before application.
This approach does not account for differences in absorption of thinning agents caused by

shading effect on cuticle and subsequent thinner absorption.  A review of the model by a 

research scientist with many years experience working on chemical thinning suggested that

only shading in the three days after application be considered, as pre-application thinning

appears to have much less impact on chemical thinner activity.   

       For night temperatures, the model uses accumulations for the 3 nights
following the application.  High night temperatures cause apple trees to
use up carbohydrate reserves at a time when depletion is not being
compensated for by photosynthesis.

       Night temperatures for each date are the average of the difference
between each hourly temperature in the following 7pm and 7am period
and 50 degrees F.  Thus, 7 out of the 12 hours (midnight to 7am)
used to calculate average night temperatures for a date actually occur
on the following calendar date.    

Range of dates included in the output table:

        The table assumes application in the morning of the stated day.  
Applications made later in the day may be more accurately assessed
by the weather influence rating for the next morning.

       The table assumes that all or most fruitlets have exceeded the 18mm 
diameter threshold that marks the end of the effective thinning window by
 482 DD43 past the McIntosh petal fall date.  

       McIntosh and synchronous cultivars probably will have reached 18mm
fruit diameter well before the last date shown on the table.  

       The later dates are shown to allow use of the table for cultivars that reach
petal fall as much as a week after McIntosh or that are slower than McIntosh
in reaching the 18mm threshold.

       The table does not address Bloom thinners because the relationships
used for assigning summary statements are based on postbloom
thinning response, and may not apply to thinning agents applied before
petal fall.

Scenarios and model statements:

        The model's summary statements only move off of the "Intermediate"
rating if there are clear indications of influence towards increased 
or decreased sensitivity.  Given the complexity and difficulty of making valid statements,

it seems better to stay neutral unless weather conditions suggest a clear tendency

for increased or decreased sensitivity to chemical thinners.

        If the temperature factor is tending toward decreased or increased 
sensitivity  to thinning, then the shading factor must be intermediate, or 
pointing in the same direction for the combined rating to be in the "Decreased 
sensitivity" or "Increased sensitivity" category.

       If the temperature factor is tending strongly toward decreased or increased
sensitivity to thinning, then the shading factor must not be pointing
in the opposite direction for the combined rating to be in the 'Stronger influence'
category.  
    

       In other words, with exceptions for very high or very cool night temperatures, the two 
factors (average night temperatures > 50F, and % shading) constrain each other.  A definitive

statement about increased or decreased sensitivity, is only possible when both factors point

in the same direction.   

Grower implementation

        The model can help growers by forecasting and highlighting weather
conditions that have characteristics associated with increased or decreased
sensitivity to thinning agents.  As such, it can be an early warning 
indicator and a planning tool to aid grower decision making.

       This could take the form of waiting for better conditions that are in the
forecast.  Or the model provide incentive to act now with the current weather
because the forecast does not offer better conditions in the next few days.

       The finalized rankings based on observed temperature and sunlight values
for the thinning application date can also be used retrospectively to help
estimate how weather might have influenced that application.  

       The estimated weather influence, combined with the record of thinning
agents used, and observed thinning results for a given set of trees, provides
some basis for identifying the ideal thinning practices for those trees in
subsequent years.

Individual weather-model statement scenarios

       The remainder of this page consists of different temperature and
shading scenarios with the summary statements they generate.  

       These scenarios are provided to demonstrate the boundary conditions
 for model assessment of postbloom weather influence on sensitivity to thinning
 agents.

Excessive sensitivity by extremely high temperatures

        An escape from the constraint requiring that high shading in order to

generate an increased thinning sensitivity rating occurs with very high night 
temperatures (category 6 or 7).

        A temperature rating of 7 (average night temperature 75F) generates
a "EXCESSIVE THINNING RISK" rating even with the lowest sunlight rating
(average sunlight of 90% or more).  To get a temperature rating of 7
requires an average night degree hour value of 75, which in turn requires
average night hour temperature of 75 F.  

        To get an average night hour temperature of 75 F over 3 nights
would typically require daytime high temperatures of 85-90 F or higher
for three days in a row.  This is well above the average daily high temperature
for the first few weeks after petal fall.  Having temperatures that high
for three days in a row shortly after bloom is unusual, but it does happen.

       A temperature rating of 6 generates a "Stronger influence for
Increased sensitivity" rating with even the lowest sunlight ranking.
(90% or more sunlight average).  To get a temperature rating of 6 requires
average night temperature of 69 F.  This would require average daytime 
high temperatures of around 84 F for three days in a row.  
This is also not typical, but becomes less unusual in the latter part
of the thinning period. 

       A temperature rating of 6 generates a "EXCESSIVE THINNING RISK"
rating  with a sunlight rating of 3 or greater (70% or less sunlight average).

Increased sensitivity by High temperatures
       If average night time temperatures rank 5 (average temperatures above
64 F), then a "Stronger influence for Increased sensitivity" can occur if the
average sunlight ranking is at least 2.8, which requires average sunlight
% over the 3-day period to be no greater than 72%.

      In other words, with average night temperature of 64 F, a "Stronger
influence for Increased sensitivity" rating  will only occur if there is
prevailing cloudy weather.

      If average night time temperatures rank 5, then an "Increased sensitivity"
rating will occur with even the lowest sunlight rating  (90% or more sunlight
average).

      In other words, with average night temperature of 64 F, an "Increased
sensitivity" rating will occur even if there is prevailing sunny weather.

Less extreme case for increased sensitivity by High temperatures
       If average night time temperatures rank 4 (average temperatures 60 F), 
then an "Increased sensitivity" rating can occur only if the average shading ranking

is at least 2.5, which requires average shading % over the 3-day period to be at least 25%.

       In other words, moderately warm night temperatures must be accompanied
by at least moderate shading to generate an "Increased sensitivity" rating.  

Increased sensitivity by High shading
       If average % shading is at its maximum rank of 5 (average value
shading above 50%, which requires very overcast conditions to get this
average over a 3-day period), then a combined rating of "Stronger
influence for Increased sensitivity" can occur if the average night temperature
rating is at least 3.9, which requires average temperatures over the
following 3 nights of at least 59.6 F.

       In other words, no matter how cloudy it is, a "Stronger influence for
Increased sensitivity" rating will only occur if average night time
temperatures are at least 59.6 F.

       If average % shading ranks 5, then an "Increased sensitivity" rating can
occur if the average night temperature rating is at least 2.8, which requires
the average temperatures over the following 3 nights to be at least 55.9 F.

       In other words, no matter how cloudy it is, an "Increased sensitivity"
rating will only occur if average night time temperatures are at least 55.9 F.

Less extreme case for increased sensitivity by High shading
       If average % shading ranks 4 (average value 40%, which is a low average

over a 3-day period), then an "Increased sensitivity" rating can occur if the

average night temperature rating is at least 3.3, which requires the average temperature

over the following 3 nights to be at least 57.4 F.

       In other words, even if it is quite cloudy over a 3-day period, an "Increased sensitivity"

rating will only occur if average night time temperatures are at least 57.4 F.

Decreased sensitivity by Very Low temperatures
        If average night time temperatures rank 0 (average night temperatures
50 F or lower), then a rating of "Stronger influence for Decreased sensitivity" 

will occur regardless of the shading rank.

Decreased sensitivity by Low temperatures
       If average night time temperatures rank 1 (average night temperatures 52 F), 
then a combined rating of "Stronger influence for Decreased sensitivity" can
occur if the average sunlight ranking is no greater than 3.3, which requires
average shading % value over the 3-day period to be 33% or more.

       In other words, if nights are cool, a "Stronger influence for Decreased sensitivity"

rating will result unless constrained by above average shading over the 3-day period.

       If average night time temperatures rank 1 (average night temperature 52 F), 
then at least a "Decreased sensitivity" will occur even at the highest shading rank.

Less extreme case for decreased sensitivity by Low temperatures
       If average night time temperatures rank 2 (average temperatures 54 F), 
then a "Decreased sensitivity" rating can occur if the average shading
rank is no greater than 3.6, which requires average shading % over the
3-day period to be no more than 36%.

       In other words, cool night temperatures will generate a "Decreased sensitivity"

rating unless combined with overcast conditions throughout the 3-day period.

Decreased sensitivity by Low shading (high sunlight)
        If average % shading ranks 1 (average value 10% or below, which
requires a string of three consecutive 'blue-sky' days), then a rating

of "Stronger influence for Decreased sensitivity" can occur if the average night

temperature rating is no greater than 2.2.  That requires average temperatures

over the following 3 nights to be no greater than 54.4 F.

       In other words, no matter how sunny it is (low shading), a rating of

"Stronger influence for Decreased sensitivity" will only occur if average night
temperatures are below 54.4 F.

       If average % shading ranks 1, then a "Decreased sensitivity" rating
can occur if the average night temperature rating is no greater than 3.2.
That requires average temperatures over the following 3 nights to be
no greater than 57.1 F.

       In other words, no matter how sunny it is (low shading), a 
"Decreased sensitivity"  rating will only occur if average night time
temperatures are below 57.1 F.

Less extreme case for decreased sensitivity by Low shading
       If average % shading ranks 2 (average value 20% for a 3-day period,
which is below average but not unusual), then a "Decreased sensitivity"
rating can occur if the average night temperature rating is no more than 2.8,
which requires average temperatures over the following 3 nights to be
no more than 55.9 F.

       In other words, under a mix of sunny & cloudy days, a
"Decreased sensitivity" rating will only occur if average night time
temperatures are below 55.9 F.