Last time | Next time |
Hopefully you've got that one in hand. It's due tomorrow, Thursday.
I want to give you some idea of direction, and let you know about some developments which, under ordinary circumstances, I might have shared in class. But now that's not quite as easy, and it's a little easier, perhaps, to do in theses chunks. I want to get you back up to speed.
So here are three things that I'd like you to do by Monday, 4/13:
This is because minima are increasing faster than maxima, so the two are getting closer together.
In some cases, it appears that maxima are not increasing much at all, perhaps because of cloudier conditions.
Minima may be increasing faster in part because of night-time warming (when it is typically cooler).
Now, for something to share with you all -- and to get you back into the game. I have some new data -- modern temperature data -- from a small town (Custar) with a good weather station not far from Bowling Green (county seat of Wood County, and the station from which most of our modern Fletcher data derived).
Here's how you might import these files into Mathematica:
Custar = Import["http://ceadserv1.nku.edu/longa//classes/mat375/days/docs/Fletcher/data/Custar.csv", "CSV"]; Custarheader = Custar[[1]] BGNormals = Import["http://ceadserv1.nku.edu/longa//classes/mat375/days/docs/Fletcher/data/BowlingGreenDailyClimateNormals.csv", "CSV"]; BGNormalsheader = BGNormals[[1]]
But you could use any software. The headers tell you what variables each data set contains.
I'd like you to find the DTR for the Custar data (you have columns for the MaxAirTemp and MinAirTemp -- the DTR would be the difference of those two variables -- DTR=max-min).
Regress the DTR, modeling it as a linear function: i.e. \[ DTR(t)=\beta t + \alpha \]
The question is whether the parameter $\beta$ is significantly different from 0, and whether in fact it's negative -- so that the DTR in Custar has been decreasing over the 38 years or so of data.
You'll need to create a variable for "year". You could just construct this as the sequence of integers from 1 to length(MaxAirTemp), then divide by 365.25 (to make up for leap years -- kind of an odd thing -- but we need to realize that years are actually 365.25 days long...!).
So carry out this experiment to see if Custar is experiencing the decrease in DTR that others are seeing around the world.
It's important that $t$ be in years for this regression -- otherwise you'd haveadd a period term, \[ Max(t) = \alpha + A\sin{\frac{2\pi(t - t_0)}{T}} \] to handle the unit conversion.
It's not hard, and you've certainly done a fair number of these. You'll perhaps be interested and surprised to see how sinusoidal the normals actually are.
Here are some Mathematica commands to get started with the normals (creating the important variables).
If there were no climate change, then we would expect the climate normals to be periodic functions -- December wrapping right back into January. However, if, as we expect, there's global warming going on, then we'll see that normals at the end of the year are actually a little higher than the normals at the beginning of the year.
And, if our hypothesis is right that the minima are changing more rapidly than the maxima, then the coefficient of the minimum will be larger than the coefficient of the maximum.
See what you discover!
Note, however, that we've had to move beyond the Fletcher data to get there....
Some good links that I might recommend (a few of which we'll focus on):
In particular, we will implement The SIR Model for Spread of Disease - The Differential Equation Model in InsightMaker.
(An excellent introduction to SIR models, from both the infectious disease and mathematical sides)
Questions:
This on-line estimator (i.e., a model!) allows one to estimate deaths, as well as death by age-category.