In March of 2015, Republican U.S. Senator Ted Cruz for the state of Texas told Seth Meyers on the show Late Night that “satellite data demonstrate for the last 17 years, there’s been zero warming. None whatsoever” (video). Sen. Cruz states that this fact is evidence that those warning of the reality and impacts of human-caused global warming are incorrect and merely alarmists.
The statistic begs the question, why 17 years? This report investigates how changing that observation period impacts the findings of whether the planet is warming.
In the above chart, all possible historical temperature trends are considered for the period of 1980-2014. The annual average temperatures and year-to-year trends are displayed in black points and dashed lines. Red and blue lines indicate whether each possible trend line results in a net increase (Warming) or no net increase (No Warming) in temperature comparing the starting and ending year annual temperatures, and shorter trends are displayed with more transparency. The specific trend referenced by Sen. Cruz is indicated with a gold, dotted line.
The data source for this analysis is the publicly available GISTEMP data. This data is maintained by NASA’s Goddard Institute for Space Studies. The global surface temperature averages are derived from NOAA meteorological stations, ERSST ocean surface temperatures, and SCAR antarctic stations. The code below can be used in R to reproduce the data for this analysis.
library(knitr); library(dplyr); library(ggplot2); library(scales)
opts_chunk$set(fig.align="center", fig.height=5, fig.width=8, fig.path="figures/")
if(!file.exists("gisdata.txt"))
download.file("http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts+dSST.txt", "gisdata.txt")
temps <- read.table("gisdata.txt", skip=8, na.strings="***", nrows=141, stringsAsFactors=FALSE)
temps <- as.data.frame(lapply(temps,as.numeric))
temps <- temps[!is.na(temps$V1),c(1,14,15)]
names(temps) <- c("year","avTempJan","avTempDec")
#convert values from relative millidegrees C to absolute degrees F
temps <- mutate_each(temps, funs(.*.018+57.2),c(avTempJan,avTempDec))
calcNetTrends <- function (temps, referenceYear) {
row.names(temps) <- as.character(temps$year)
temps[paste0("xYearTrendJan",referenceYear)] <-
temps[referenceYear,"avTempJan"] - temps$avTempJan
temps[paste0("xYearTrendDec",referenceYear)] <-
temps[referenceYear,"avTempDec"] - temps$avTempDec
temps[paste0("numYears",referenceYear)] <-
temps[referenceYear,"year"] - temps$year
temps
}
#Calculate various temperature trends to compare with Cruz's stat
temps <- calcNetTrends(temps, referenceYear="2013")
temps <- calcNetTrends(temps, referenceYear="2014")
#Caluclare all possible past trendlines for the main chart
select(na.omit(temps), year, avTempDec) %>% merge(., ., by=NULL) %>% filter(year.y<year.x) %>% mutate(trend=factor(avTempDec.x - avTempDec.y <= 0, labels=c("Warming", "No Warming"))) -> segs
#Set parameters for the main chart
start_year <- 1980
yscale <- c(57, 58.5)
ascale <- c(.2,.6)
First, Cruz’s claim that there is no net warming for 17 years from 1998 - 2013 will be examined.
Years from 2013 | °F warming in 2013 (Jan-Dec) | °F warming in 2013 (Dec-Nov) | Years from 2014 | °F warming in 2014 (Jan-Dec) | °F warming in 2014 (Dec-Nov) | |
---|---|---|---|---|---|---|
1996 | 17 | 0.504 | 0.504 | 18 | 0.630 | 0.630 |
1997 | 16 | 0.270 | 0.270 | 17 | 0.396 | 0.396 |
1998 | 15 | -0.018 | -0.036 | 16 | 0.108 | 0.090 |
1999 | 14 | 0.360 | 0.324 | 15 | 0.486 | 0.450 |
2000 | 13 | 0.360 | 0.306 | 14 | 0.486 | 0.432 |
In this table both 2013 and 2014 are considered as the comparison year as well as whether the annual temperature average is based on the calendar year (January to December) or the seasonal year (December to November). In no case is Sen. Cruz’s claim of 17 years without warming strictly accurate. However, it appears that he was referring the 15 year period from 1998-2013, which began 17 years prior to the time of his statement and does indeed indicate a lack of warming. To find this warming trend, however, he ignored the temperature data for 2014, even though it was available at the time of his statement. As the seasonal year data is slightly more favorable to Sen. Cruz’s claims than the calendar year data, it will be used throughout this report.
maps<-matrix(c("black", "tomato3","deepskyblue2", "gold", "dashed", "solid", "solid", "dotted", "Annual Temperature","Trend with Warming", "Trend without Warming", "Ted Cruz's Trend"), nrow=4, dimnames=list(c("Temp","Warming", "No Warming", "Cruz"),c("color","linetype","label")))
ggplot(data=temps[temps$year>=start_year,], aes(x=year, y=avTempDec)) +
geom_segment(aes(x=year.x, xend=year.y, y=avTempDec.x, yend=avTempDec.y, color=trend, alpha=(year.x-year.y), linetype=trend), data=segs[segs$year.y>=start_year,], size=1.5) +
geom_line(aes(linetype="Temp", color="Temp"), size=.5) +
geom_segment(aes(x=1998, xend=2013, y=temps[temps$year==1998,"avTempDec"],yend=temps[temps$year==2013,"avTempDec"], color="Cruz", linetype="Cruz"), size=1) +
geom_point(linetype=2,size=2) + labs(title=paste0("Temperature Trendlines ",as.character(start_year),"-2014"), x="Year",y="Global Average Surface Temp. (°F)") +
scale_y_continuous(limits=yscale) + scale_alpha_continuous(range=ascale) +
scale_color_manual(name="leg", values=maps[,"color"], breaks=rownames(maps), labels=unname(maps[,"label"])) +
scale_linetype_manual(name="leg", values=maps[,"linetype"], breaks=rownames(maps), labels=unname(maps[,"label"])) +
guides(alpha=FALSE, linetype=guide_legend(title=NULL, override.aes=list(size=.8)), color=guide_legend(title=NULL)) +
theme(title=element_text(size=20), axis.title.y=element_text(size=14, vjust=1), axis.title.x=element_text(size=14), axis.text=element_text(size=12), legend.text=element_text(size=10), legend.position="bottom")
It is already clear that the specific period of 1998-2013 is not representative of overall global temperature trends, especially considering that 1998 is known for its strong El Nino Southern Oscillation that was described at the time as a “record breaking event” with “unusual extremes” by the National Climactic Data Center. However, the period charted above does not portray the entire temperature record. Below, the full record from 1880-2014 an all 8911 possible trends are charted with the same methodology.
As in the first plot, the annual average temperatures and year-to-year trends are displayed in black points and dashed lines. Red and blue lines indicate whether each possible trend line results in a net increase (Warming) or no net increase (No Warming) in temperature comparing the starting and ending year annual temperatures, and shorter trends are displayed with more transparency. The specific trend referenced by Sen. Cruz is again indicated with a gold, dotted line. However, this plot uses all available GISTEMP data, dating back to 1880.
The claim that there has been no warming for 17 years is false. The net change in average global temperature for the 15 year period from 1998 to 2013 is a decrease, but the vast majority of trends and the overall trend clearly indicate that the planet is warming.