Mini-Project 3
We have been asked by Togo's Direction Generale de la Meteorologie
Nationale to help determine if Togo's surface temperatures are
increasing (or changing, and how), in light of the predictions by climate scientists.
- In Mini-Project 1, you worked with annual data for a particular city, and produced a report. Those reports are contained here.
- In Mini-Project 2, you worked with monthly data for a different city, with a different partner, and produced a report. Those reports are contained here.
- In Mini-Project 3, you worked with monthly rainfall data for a third city, with a different partner, and produced a report. Those reports are contained here.
The Togolese have now also provided us with monthly rainfall
data on those same cities: here
are those data (as provided to me).
We would like to use that data to
- determine if there is a relationship between rainfall and temperatures.
- create a model (including periodicity) of rainfall that
would provide them with predictions for the next
several years into the future.
- You might experiment with more complicated periodic functions. For
example, incorporate pairs of sine and cosine terms with other
periods: for example
\[
\sin(\frac{2\pi t}{2})
\]
has a period of two years;
\[
\sin(\frac{2\pi t}{1/2}) = \sin(4 \pi t)
\]
has a period of half a year. You get the picture. I might be tempted to
try some periods of the form
\[
\sin(2^n\pi t)
\]
where $n$ is an integer (positive and negative!). There's a limit (the
Nyquist limit!) to what you can "scope out". You might have a look at
that.
Each of you has been assigned a particular city, different from the one you
studied in the first mini-project. You should all be working with a different
partner (or partners). Group number corresponds to the city as initially given:
1 Lome
2 Tabligbo
3 Kouma-Konda
4 Atakpame
5 Sotouboua
6 Sokode
7 Kara
8 Niamtougou
9 Mango
10 Dapaong
Here are your groups:
Last Name,First Name,Group
Hardesty,Austin,1
Odhiambo,Donna,1
Gillespie,Leah,2
Driehaus,Rachel,2
Campbell,Austin,3
Englert,Jacob,3
Zembrodt,Alli,4
Compton,Lizzy,4
Edwards,Connor,5
Dufek,Sally,5
Ruwe,Maria,6
Ficke,Terra,6
Koors,Jacob,7
Farmer,Alyssa,7
May,Adam,8
Frink,Clayton,8
McMahon,Maria,9
Gall,Matthew,9
Nielsen,Patrick,10
Milesky,Chris,10
- You will begin with an evaluation phase, due 2/23, which
you hand to me (one to two pages, typed) and which is
strictly anonymous (i.e. no one else will ever see
it):
- Evaluate your own report, especially comparing it to other reports.
- Compare your partner(s)' and your contributions to the
report. Perhaps you took different roles, which you can describe.
- Evaluate the report by the second group to study your new
city. Identify things that you like, and things that
you think could have been done better.
- Your group should produce a typed report in pdf
format (no more than ten pages, including figures), as
well as any scripts or code you used to produce your
report.
For this report, you will do the following:
- The first and second mini-project reports are preliminary
work. You can cite them, discuss them, and otherwise use them
as you wish. You can suggest improvements, corrections, or pin
gold medals on them. It's just like all published research --
there for people to use, appreciate, or take pot shots at!:)
You may use them as the basis for your new report, but credit
given where credit is due.
Your report will extend those reports -- that was preliminary
work. This is similar work, with a different purpose.
- Include background information on your city (or region), focusing especially
on anything that might impact climate.
- Continue to ruthlessly identify outliers in the
rainfall data. If anything smells at all suspicious,
you should identify it, create a record of your
suspicions, and we should ask the question of the
Togolese.
- Determine if rainfall for your city varies with temperature over
time. Discuss any characteristics of the data that seem
relevant.
- Provide and discuss your best model of $rainfall(time,
temperature)$. Evaluate your models.
- Provide graphs of the data with their model(s), with
labels and title.
- Provide graphs of the residuals, with labels and title.
- Discuss what additional information might be useful moving
forward, problems with your data, and any information you'd
like to know from the Togolese meteorologists.
- Submit your report electronically, as well as a paper copy. Provide
supporting scripts, etc., electronically only.
The report is due Friday, 3/2.
Website maintained by Andy Long.
Comments appreciated.