How hot has it been in Victoria?

Note:The BOM uses the mean of tem­per­a­tures in the peri­od 1961–1990 as the ‘base­line’ for anom­alies. The NASA data from RSS cal­cu­lates anom­alies as dif­fer­ences from the mean in the peri­od 1979–2008 (for the rel­e­vant series). But in this case, the dif­fer­ent base­lines don’t matter.I’m inter­est­ed only in whether the BOM claim of a huge anom­aly (> 1° cel­sius) is plau­si­ble. It sure doesn’t look that way from space.

The smooth­ing in the graph is based on a sim­ple, two-sided, mov­ing aver­age over 13 months. It’s unnec­es­sary; only a piece of eye-can­dy.

Here’s the R script used to cre­ate the graph. It should be able to han­dle future updates to the RSS data. Please note that you’ll need the “Hmisc” library from CRAN for the last step in the script.

Dis­own­er: I’m an R ama­teur, so…this is a kludge, your milage may vary etc. etc.

#
# Load S-hemisphere (-70s to -20s) land anomalies
# Smooth them using a two-sided filter (13 years)
# Plot the anomaly time-series and the smoothed series
#
# for the tick marks on the graph
library(Hmisc)
#
# get the data: land temperature anomalies
file<-"ftp://ftp.ssmi.com/msu/monthly_time_series/rss_monthly_msu_amsu_channel_tlt_anomalies_land_v03_2.txt"
f<-read.table(file, skip=3, fill=TRUE)
# create a time-series
# f$V6 is the vector corresponding to data for 70s - 20s latitude
# Substitute: f$V3 = global; f$V4 = tropics; f$V5 = nth hemisphere
# f$V7 = nth temperate; f$V8 = sth temperate
Anomaly=ts(f$V6,start=f$V1[1], frequency=12)
#
# a vector of weights
k<-c(0.0025,0.005,0.1,0.25,0.5,1,1,1,0.5,0.25,0.1,0.005,0.0025)
# smooth the anomaly series
A_Smooth<-filter(Anomaly,(k/sum(k))) # filter the anomalies
#
# the date of the data
my_months<-c("January","February","March","April","May","June","July","August","September","October","November","December")
#
# now plot the data
plot(Anomaly, col="darkseagreen", lty=1, xlab="",ylab="degrees c", main="Southern Hemisphere Lower Troposphere (Land) Anomaly", sub=paste("Source: Remote Sensing Systems, to",
my_months[f$V2[length(f$V2)]], f$V1[length(f$V1)], sep=" "), cex.sub=0.7, font.sub=3, cex.axis=0.8)
# overplot the smoothed series
lines(A_Smooth, lwd=3,col="darkorange")
# add a legend
legend("bottomright",c("anomaly","smoothed"), cex=0.8, lwd=2,col=c("darkseagreen","darkorange"),horiz=TRUE, bty="n", inset=0.02)
# and a 'mid-line' to make the anomaly easier to read
abline(h=0.0,lty=2,col="darkslategray4")
# some tick marks
minor.tick(nx=5)
#
#
#

No Comments

Leave a Reply

Your email is never shared.Required fields are marked *