Climate Change

This data file contains the monthly average lower troposphere temperature since 1979 according to Remote Sensing Systems (RSS). The data are obtained by the National Oceanographic and Atmospheric Administration (NOAA) TIROS-N satellite, and interpreted by Dr. Carl Mears (RSS). After downloading the data to your R's work space, use the following command to load it to R.

temp <- read.csv("RSS_Monthly_T_anonalies.csv", comment.char="#")

The option comment.char="#" tells R to ignore lines that begin with the "#" character. The data frame has 3 columns. The first column is year; the second column is month and the third column is temperature anomaly in °C. The term temperature anomaly means the departure from a reference temperature value.

a. To analyze the data, we first want to convert 'year' and 'month' to a specific time. We can assign a date corresponding to the middle of the month and use R's Date class to store the date. For example, March 2010 can be converted to the date "2010-03-15". Which of the following commands can be used to create a new column 'date' in the data frame temp that store the date corresponding to the middle of the month?
temp$date <- as.Date(paste0('year', '-', 'month', '-15'))
temp$date <- as.Date(temp$year, '-', temp$month, '-15')
temp$date <- as.Date(paste0(temp$year, '-', temp$month, '-15'))
none of the above

You can make a plot of temperature anomaly vs time using the command

plot(T_anomaly ~ date, type='l',data=temp, ylab="Temperature Anomaly (deg C)",xlab="Year", las=1)

The option type='l' tells R to join the points to a line. You should see a plot similar to the following.

b. Fit a linear model predicting the temperature anomaly from date using the following command

fit <- lm(T_anomaly ~ date, data=temp)

In the Date class, dates are stored internally as the number of days since 1970-01-01. As a result, the slope of the regression represents the change of temperature per day. A positive slope means global warming. Enter the slope to 3 significant figures. You can use the e-notation for numerical answer. For example, you can enter 1.23e-4 instead of 0.000123.

Rate of temperature change = °C/day

c. Convert the slope in part (c) to °C/decade, using 1 decade = 3652.5 days. Enter your answer to 3 decimal places.

Rate of temperature change = °C/decade