In attendance will be participants from government, research and meteorological institutes, universities, and other key partners. They will mostly come from the South Asian (Indo-Gangetic Plains) region, but West and East Africa will be represented as well. The workshop will provide overviews of and training in climate modeling techniques and tools, including downscaling, weather generators, and modules. It will feature some CCAFS-produced or -refined outputs like MarkSIM and the climate analogue tool. The focus will be on how to use these instruments in real-world applications, so case studies will also be explored.

Following the first session about mapping with R I would like to propose you two exercises:

1) Plot a map of the world that presents information on the number of earthquakes around the world (by country)

Hint: The Significant Earthquake Database contains information on destructive earthquakes from 2150 B.C. to the present

2) Plot a map of Spain that shows any socio-economic variable by province (eg. population)

Hint:  You can find Spanish map files and data at Instituto Nacional de Estadística.

2. Data mining

After reading Development of Arab Water Sustainability Index Using Principal Component Analysis and based on what we have seen in class about Principal Component Analysis and Clustering, try to analyse relevant water resources data for Middle East countries. You can find this data from the FAO’s Information System on Water and Agriculture.

3. Meta-analysis

Nonulcer dyspepsia is characterized by a variety of upper abdominal symptoms in
the absence of an organic disease. Due to the high frequency of these symptoms an
empirical probative therapy of nonulcer dyspepsiawithout prior diagnostic procedures
has been recommended for the management of these patients. For this empirical
therapy either acid blocking substances such as histamine H2-receptor antagonists or
gastroprokinetics such as cisapride or domperidone have been suggested.
A major problem in treatment studies of nonulcer dyspepsia is the relatively high
and variable rate of placebo responders. The reason for this variability is unclear, but
based on this placebo response rate, it becomes difficult to draw valid conclusions on
the treatment recommendations. This highly variable response rate can be overcome
if available clinical studies are put together and evaluated with appropriate statistical
methods in a meta-analysis. Only those placebo-controlled studies were included
which used similar criteria for “treatment success” andwhich documented their results
sufficiently for a possible reanalysis.

The data reported here are the number of successes and the number of patients in the treatment group aswell as in the control group.

Hartung, J.andKnapp, G. (2001b)Arefinedmethod for themeta-analysis of controlled clinical
trials with binary outcome. Statistics in Medicine, 20, 3875–3889. (7, 8, 9, 18)

• is it safe?
• is it clean?
• is it cheap?
• is it fair?

You may find inspiration for a discussed analysis in the Statistical Analysis and Visualization of the Drug War in Mexico by Diego Valle.

Some useful links:

Nuclear power plant accidents: listed and ranked since 1952

Nuclear energy data

The Nuclear Energy Agency’s annual brown book