As KSSV said, the random function generates data very nicely. But this distribution might need to be adapted to your needs.
To adjust to your problem you need to add the desired mean of your data and multiply the distribution.
mean_temperature = 15*sin(0:0.1:6)+15;
random_variation = rand(1, 61)*20-10;
plot(mean_temperature+random_variation)
And if you want a non uniform distribution, you need to multiply the random_variation with a function instead of the value 20.
hope this helps you get started
