Table 2.

Quantile regression results for maximum temperature change over time (years), Simpson Desert, central Australia

Weather station Quantile	Estimate	CI	t-value	P
Boulia
0.05	0.01	0.003–0.01	1.62	0.11
0.1	0.007	0.002–0.01	1.65	0.10
0.5	0.003	−0.003–0.005	0.98	0.23
0.9	0.005	−0.003–0.007	0.99	0.32
0.95	−0.002	−0.005–0.004	−0.31	0.76
Alice Springs
0.05	0.002	−0.02–0.02	0.28	0.78
0.1	0.003	−0.01–0.01	0.48	0.63
0.5	0.004	−0.001–0.009	1.29	0.20
0.9	−0.002	−0.004–0.001	−0.61	0.54
0.95	−0.003	−0.004–0.002	−1.07	0.29
Birdsville
0.05	0.003	−0.004–0.06	0.12	0.90
0.1	0.04	0.01–0.06	1.95	0.06
0.5	0.03	0.02–0.04	4.30	<<0.001*
0.9	0.02	0.01–0.05	1.55	0.12
0.95	0.03	0.006–0.04	2.73	0.01*
Oodnadatta
0.05	0	−0.004–0.03	0	1
0.1	0.01	−0.001–0.02	1.97	0.05*
0.5	0.02	0.01–0.03	3.98	<<0.001*
0.9	0.02	0.01–0.03	2.99	0.004*
0.95	0.03	−0.002–0.04	3.50	<<0.001*

In each row a significant P-value indicated by * implies that there is a nonzero slope for that quantile. For each weather station (Boulia, Alice Springs, Birdsville, Oodnadatta) the slope estimate for each of the extreme quantiles (0.5, 0.1, 0.9, and 0.95) was compared with the median quantile (0.5). Overlaps in the confidence intervals (CI) indicated similar rates of change for all comparisons.