Statistical climatology
(Neptun code: metstatkl0g17em)
Main topics:
1. Basic definitions and concepts of probability theory and mathematical statistics
2. Overview of discrete and continuous probability distributions
3. Methods to estimate the parameters of probability distributions
4. Statistical hypothesis testing
5. Overview of stochastic processes (i.e., linear time series models)
Course outlines:
The application of statistical methods on samples obtained from meteorological measurements often requires to estimate the probability distribution of the underlying random variables. Consequently, at first, the basic concepts and definitions of probability theory and mathematical statistics will be overviewed e.g., real-valued random variables, probability distribution function (pdf), density function, theoretical moments, statistical relationships (covariance, correlation), and empirical cumulative distribution function (ecdf), (relative) frequency, histogram. Then, some of the most frequently used probability distributions are discussed. Probability distributions (e.g., Poisson, normal, Weibull) will be overviewed from which the meteorological samples may come from. The parameters of the probability distribution can be estimated based on the sample. For that purpose, the method of moments will be described. After that, the histogram/ecdf created from the sample is compared to the density function/pdf estimated from the sample. The goodness of the estimation can be verified by hypothesis testing which is also discussed. Finally, we examine random variables that depend on time, therefore, we move on to the topic of stochastic processes. The definitions of some notable stochastic processes — i.e., linear time series models such as autoregressive models — are described. During the semester, the topics mentioned above are illustrated with meteorological examples.
Additional information:
The course concludes with a written exam. The semester consists of 13 (90 minutes long) lectures. For students who receive Stipendium Hungaricum scholarship, lectures are hold on Wednesday from 10:15 to 11:45 (Northern Block, room 6.129). Until 30th March the first two topics listed above will be discussed and 8 lectures will be held. Guest students may also participate on the lectures on Wednesday, however, repetition of the first two topics is possible on every Tuesday from 16:00 to 17:30 (Northern Block, room 6.129).
Statistical Methods for Climate Data Analysis
(Neptun code: meteghido0g17em)
The course is basically an R programming language course which has two versions: one for beginners and one for advanced learners.
A short introduction to R can be found here
Updated on 21.02.2022
Main topics for beginners:
1. Basics of the R programming language
2. Reading text files into R
3. Plotting data in R on x-y plots (creating timeplots)
4. Handling multidimensional arrays in R, plotting data fields (creating simple maps)
For beginners, the main goal of this course is to learn the basics of the R programming language to handle data tables and multidimensional arrays (e.g., txt, csv, and netCDF files) in R. Arithmetic operations will be applied on the data. Time series will be visualized on x-y plots, while simple maps - with Cartesian projection - will be created from data fields.
Main topics for adavanced learners:
1. Effective ways to handle data tables in R (filtering and gathering data)
2. Analysis of the distributions of climate data
3. Plotting the results on various chart types
Course outlines:
For advanced learners, some topics of data science will be presented. In the course, tables of climate data will be handled in R (e.g., we will filter and join tables according to different aspects). Then basic statistical analyses on the climate data will be applied concerning their distributions. The results will be plotted on a variety of chart types (e.g., on histograms, density plots, quantile-quantile plots and barplots). Although we process climate data, our goal is to gain knowledge that will allow us to perform data analysis on other research fields. This version of the course requires basic R and statistical knowledges. The course concludes with an oral presentation.
Additional information:
The beginners' course concludes with a written exam, while the advanced learners' course concludes with an oral presentation. The semester consists of 6 (90-100 minutes long) lectures. For students who receive Stipendium Hungaricum scholarship the course ends on 23th March. Guest students can start the course on 30th March or on 6th April from the first topic listed above. Lectures will be held on every Wednesday from 8:30 to 10:00 or from 12:30 to 14:00 (Northern Block, room 6.132).
Last updated on 20.03.2022