Jacqueline Wingen: Manual chill

This chapter explains how to calculate Chilling Hours using R and the chillR package. Chilling Hours measure the number of hours where temperatures are between 0°C and 7.2°C, which is important for certain plants to meet their cold requirements during dormancy and grow properly.

Data Requirements

The calculation requires hourly temperature data, which is not always available. The chillR package provides tools to approximate data from daily records. In this example, the dataset Winters_hours_gaps is used, containing hourly temperature data recorded in 2008 from a walnut orchard in Winters, California.

Data Preparation

The chillR package is loaded using library(chillR). The relevant columns (year, month, day, hour, temperature) are extracted and stored in a new dataset named hourtemps, ensuring the correct format for calculating Chilling Hours.

hourtemps <- Winters_hours_gaps[,c("Year",
                                   "Month",
                                   "Day",
                                   "Hour",
                                   "Temp")]

Manual Calculation of Chilling Hours

Chilling Hours are defined as any hour where the temperature falls between 0°C and 7.2°C. In R, this is implemented using a logical condition:

hourtemps[, "Chilling_Hour"] <- hourtemps$Temp >= 0 & hourtemps$Temp <= 7.2

A new column Chilling_Hour (TRUE/FALSE) is created to indicate whether a given hour qualifies. The total number of Chilling Hours can then be calculated using sum(hourtemps$Chilling_Hour).

Automation with Functions

To simplify the process, a function CH() was created to automatically add the Chilling_Hour column:

CH <- function(hourtemps) {
  hourtemps[, "Chilling_Hour"] <- hourtemps$Temp >= 0 & hourtemps$Temp <= 7.2
  return(hourtemps)
}

Additionally, a function sum_CH() was developed to calculate the total number of Chilling Hours between two specific dates:

sum_CH <- function(hourtemps, Start_Year, Start_Month, Start_Day, Start_Hour, 
                              End_Year, End_Month, End_Day, End_Hour) {
  hourtemps[,"Chilling_Hour"] <- hourtemps$Temp >= 0 & hourtemps$Temp <= 7.2

  Start_Index <- which(hourtemps$Year == Start_Year & hourtemps$Month == Start_Month &
                       hourtemps$Day == Start_Day & hourtemps$Hour == Start_Hour)
  End_Index <- which(hourtemps$Year == End_Year & hourtemps$Month == End_Month &
                     hourtemps$Day == End_Day & hourtemps$Hour == End_Hour)

  CHs <- sum(hourtemps$Chilling_Hour[Start_Index:End_Index])
  return(CHs)
}

This function uses the which() function to identify the relevant rows in the dataset based on the selected time range.

Optimization of Function Parameters

Instead of passing year, month, day, and hour separately, compact strings in the format YEARMODAHO (e.g., 2008040100 for April 1, 2008, at 00:00) can be used. The function extracts values using substr() and converts them into numeric values.

sum_CH <- function(hourtemps, startYEARMODAHO, endYEARMODAHO) {
  hourtemps[, "Chilling_Hour"] <- hourtemps$Temp >= 0 & hourtemps$Temp <= 7.2

  startYear <- as.numeric(substr(startYEARMODAHO, 1, 4))
  startMonth <- as.numeric(substr(startYEARMODAHO, 5, 6))
  startDay <- as.numeric(substr(startYEARMODAHO, 7, 8))
  startHour <- as.numeric(substr(startYEARMODAHO, 9, 10))

  endYear <- as.numeric(substr(endYEARMODAHO, 1, 4))
  endMonth <- as.numeric(substr(endYEARMODAHO, 5, 6))
  endDay <- as.numeric(substr(endYEARMODAHO, 7, 8))
  endHour <- as.numeric(substr(endYEARMODAHO, 9, 10))

  Start_Index <- which(hourtemps$Year == startYear & hourtemps$Month == startMonth &
                       hourtemps$Day == startDay & hourtemps$Hour == startHour)
  End_Index <- which(hourtemps$Year == endYear & hourtemps$Month == endMonth &
                     hourtemps$Day == endDay & hourtemps$Hour == endHour)

  CHs <- sum(hourtemps$Chilling_Hour[Start_Index:End_Index])
  return(CHs)
}

Application Example

Using the function sum_CH(), it was calculated that between April 1st and October 11th, 2008, the walnut orchard experienced 77 Chilling Hours:

sum_CH(hourtemps, startYEARMODAHO = 2008040100, endYEARMODAHO = 2008101100)

[1] 77

`Exercises` on basic chill modeling

Write a basic function that calculates warm hours (>25°C).

WH <- function(data)
  {data[, "Warm_Hour"] <- data$Temp > 25
  return(data)
}

Apply this function to the Winters_hours_gaps dataset.

WH(Winters_hours_gaps)

Year	Month	Day	Hour	Temp_gaps	Temp	Warm_Hour
2008	3	3	10	15.127	15.127	FALSE
2008	3	3	11	17.153	17.153	FALSE
2008	3	3	12	18.699	18.699	FALSE
2008	3	3	13	18.699	18.699	FALSE
2008	3	3	14	18.842	18.842	FALSE
2008	3	3	15	19.508	19.508	FALSE
2008	3	3	16	19.318	19.318	FALSE
2008	3	3	17	17.701	17.701	FALSE
2008	3	3	18	15.414	15.414	FALSE
2008	3	3	19	12.727	12.727	FALSE

Extend this function, so that it can take start and end dates as inputs and sums up warm hours between these dates.

sum_WH <- function(data, 
                   startYEARMODAHO,
                   endYEARMODAHO)
  
{data[,"Warm_Hour"] <- data$Temp > 25

startYear <- as.numeric(substr(startYEARMODAHO, 1, 4))
startMonth <- as.numeric(substr(startYEARMODAHO, 5, 6))
startDay <- as.numeric(substr(startYEARMODAHO, 7, 8))
startHour <- as.numeric(substr(startYEARMODAHO, 9, 10))

endYear <- as.numeric(substr(endYEARMODAHO, 1, 4))
endMonth <- as.numeric(substr(endYEARMODAHO, 5, 6))
endDay <- as.numeric(substr(endYEARMODAHO, 7, 8))
endHour <- as.numeric(substr(endYEARMODAHO, 9, 10))


Start_Date <- which(data$Year == startYear &
                    data$Month == startMonth &
                    data$Day == startDay &
                    data$Hour == startHour)

End_Date <- which(data$Year == endYear &
                  data$Month == endMonth &
                  data$Day == endDay &
                  data$Hour == endHour)

WHs <- sum(data$Warm_Hour[Start_Date:End_Date])
return(WHs)
}

Application Example:

sum_WH(Winters_hours_gaps, startYEARMODAHO = 2008080100, 
                           endYEARMODAHO = 2008083100)

[1] 283

During the month of August 2008, from the 1st to the 31st, the walnut orchard experienced a total of 283 warm hours (defined as hours when the temperature exceeded 25°C).

Manual chill