Sie sind hier: Startseite Studiengang Profillinienmodule Environmental Modelling and GIS

Environmental Modelling and GIS

We are living in an age of data: sensors, satellites and stations record environmental states around the world, in many systems. The profile "Environmental modelling and GIS" aims to provide the technical skills that are needed to handle data, both in terms of collecting and storing, as well as in analysing and visualising them. Additionally, we reflect on the use of these data on the background of specific environmental systems (soil, forest, etc), so that a critical attitude towards which questions environmental data can and cannot address. One running theme of this profile is the upscaling of information, e.g. from many local measurements to a landscape-scale representation, and the prediction to new locations and environmental conditions.

The profile clearly targets students with a knack for computer-work. Basic knowledge of GIS and R are prerequisites, and we expect the willingness to invest into learning new software and languages (SQL, Python). Modules differ in their focus, requiring students to familiarise themselves with the topic and the indicated prerequisites before each module.

We expect

* basic knowledge of working with R (at the level presented in "R for Beginners")

* basic knowledge of GIS and typical spatial data structures (as taught e.g. in the BSc-course geomatic)

* basic knowledge of SQL as data base query language (see a http://pickatutorial.com/ for introductory tutorials)

 

Coordinator of the Elective Track

dormann.jpg 

  

 

 

Prof. Dr. Carsten Dormann

Department of Biometry and Environmental System Analysis

Phone: +49 761 203-3749

5 reasons for the Elective Track "Environmental Modelling and GIS"

     * will follow soon*

Employment options

Data skills are under increasing demand in administration, academics, NGOs and industry. Environmental applications become more sensor-driven, with satellite pictures, continuous weather and air quality sensors, and large surveys and inventories challenging the classically educated environmental scientist. The profile "Environmental Modelling and GIS" provides the background in data handlingstatistical analysis and visualisation to make its graduates highly attractive to employers in all market sectors. Our alumni analysis demonstrates a full uptake by the market, with a substantial proportion of graduates now working in data base analysis and web-based data services for environmental data.

Programme overview 

EMG Studienverlauf

 

Detailed module descriptions

**können zurzeit aufgrund der Corona Pandemie abweichen**

First semester (winter)

Data Collection, -Storage, -Management 

Module coordination

Dr. M. Stölzle

Additional lecturers

M. Mälicke

Teaching and learning methods

Lectures, fieldwork, practical computing exercises

Type of examination

Two-part Portfolio (Time-Series Data + Spatial Information)

Requierments

  • Basic knowledge about statistics and GIS
  • Knowledge and beginner experience with R, R Studio (e.g. load/write data files, simple plots, data types)
  • Experience with the GUI of R Studio!

Syllabus

In this module, you will receive the basics for the acquisition, handling and processing of space-time data, to a level suitable for a later modelling. The frame topic will be the experimental description of the ‘urban heat island effect‘ in the City of Freiburg. Participants will document all work steps in a portfolio. The course used a GitHub framework to work together on two different topics and submit exercises:

Time-Series Data: Analog and digital methods of data acquisition in the field are presented and discussed. This extends from the basic elements of analog field protocols (field book) to complex data logging. The students themselves will program temperature data loggers, install them in their place of residence, read the recorded data, and then critically check the accuracy of this data. For comparison, time-series data is downloaded from the internet. All time series are subjected to a comprehensive quality control procedure in R. Errors in the time series are deleted, and the resulting data gaps are filled using various methods. As a result characteristic parameters can be determined for the temperature profile. One important component of the course is data visualization (e.g. maps). You will learn about different data types, theory of data visualization and how visualization can be done in R and GIS (best practice guide). You will also work with climate data and climate indices on a larger scale (Baden-Württemberg) to combine temporal and spatial data analysis.

Databases with Spatial Data: Using Q-GIS, the parameters will be spatially interpolated and compared with existing meta data of the city (e.g. building density).  Particular emphasis is placed on transferring data between ‘R‘ and Q-GIS. This is followed by an introduction to SQL and common database systems. A simple storage solution will be presented and used to persist data, run analyses on the data collected and comparisons to the data of previous years

Learning goals and qualifications

Students will

  • know the basics of data collection in the field via modern digital methods
  • know data sources, data types, fundamental data formats, data visualization
  • be able to independently collect data on the ground and use Internet data sources
  • be able to import collected data into data management software and data quality control of time series
  • be able to do spatial interpolation of time series data and evaluation of their accuracy

Literature

Zahumenský, I (2004): Guidelines on Quality Control Procedures for Data from Automatic Weather Stations, WMO, Geneva.

Numeric Modelling of Processes 

Module coordination

PD Dr. Helmer Schack-Kirchner

Additional Lecturers

N.N.

Teaching and Learning Methods

Independent computer-based training

Type of Examination

Written exam (32%) and homework on a model problem (68%)

Requierments

  • Basic knowledge in chemistry, physics and biology of soils
  • Confident use of “R”

Syllabus

The heterogeneity and interdependence of processes hamper the provision of hard figures describing the state and development of ecosystems. On the other hand, many laboratory studies and local observation series exist concerning only single aspects of ecosystems. A well-known means to combine basic laws and specific observations are numerical process models. In these models, isolated sections of the system are simulated with mathematical methods, results are compared to the observations and hypotheses are tested. Within this module, students are trained in the basic skills to develop such models, as well as their critical applications. 

The principal teaching object of the module is an integrative chemical-physical-biological model to describe the dynamics of carbon dioxide in soil systems including production by biota, gaseous and dissolved transport, and chemical speciation including pH effects and rock dissolution. The model includes empirical fitting, ordinary and partial differential equations and interactive processes. Complementary aspects of numerical models are trained with exemplary problems outside the main model, such as soil compaction, random-walk problems or cellular automata.

During the training time, the participants mostly develop their own sections of code that are then assembled into a final program. R-statistics is used as the programming platform. The following programming skills (not complete) are covered: Data types, modularization, user-defined functions, code organization, matrix-type vs. loop processing, package ODEsolve including method-of-lines for partial differential equations.

Learning goals and qualifications

Students will

  • be able to apply basic programming techniques
  • know of fundamental steps in modelling
  • be able to transfer simple environmental processes into differential equations
  • be able to program environmentally relevant process models
  • be able to evaluate existing models with respect to their scientific value

Reading

Soetaert, K., & Herman, P. M. (2008). A practical guide to ecological modelling: using R as a simulation platform. Springer Science & Business Media.

 

Advanced Statistics - Mixed Effect Models with R 

Module coordination

Dr. Arne Schröder

Additional lecturers

Severin Hauenstein, Prof. Dr. Carsten Dormann

Teaching and learning methods

Lectures, demonstrations, practical work and exercises

Type of examination

Written computer-based exam (3,5h), presentation of exercise results (Studienleistung, not graded)

Syllabus

The module teaches competences for the development (research) and application (practice) of advanced but important statistical models in the environmental sciences.

The module focuses on mixed effects models and their application in R. Mixed effects models are powerful tools to deal with structure and heterogeneity in environmental data arising from such common practices as multiple sampling of units, grouping units at various hierarchical levels, or spatial sampling. A rough estimation shows that 80-90 % of environmental studies require mixed effects models to analyse their data. However, mixed effects models are also complex and sometimes difficult to apply and interpret. More, they are developing fast and their possibilities expand continuously. The module´s goal is to teach students the basics of mixed effects models on which to build on when analysing their own data. The course thus extends statistical knowledge and its application as conveyed by other courses at the faculty. Topics covered will be repeated measurement ANOVA, generalised least squares (GLS), linear mixed models (LMMs), generalised linear mixed models (GLMM) and possibly generalised additive mixed models (GAMM).
All topics will be taught in the free software R, mainly using the R-packages nlme, lme4, gls, aov and their add-on

Learning goals and qualifications

Students will

  • be able to apply and interprete mixed effects models
  • be able to solve complex statistical tasks independently using the software R and its relevant resources
  • get competences for the development (research) and application (practice) of advanced but important statistical models in the environmental sciences.

Prerequisites

Basic statistical knowledge in statistics (ANOVA, ANCOVA, GLMs, GAMs) and R

Reading

  • *Paradis, E. R for Beginners (https://cran.r-project.org/doc/contrib/Paradis-rdebuts_en.pdf)
  • Crawley M (2007) The R Book. Wiley.
  • Zuur A et al. (2007) Mixed Effect Models and their Extensions in Ecology with R. Springer.
  • Bolker B et al. (2009) Generalized linear mixed models: a practical guide for ecology and evolution. Trends in Ecology and Evolution 24:127 – 135. https://www.sciencedirect.com/science/article/pii/S0169534709000196
  • *R-Dokumentation unter http://cran.r-project.org/other-docs.html, etwa http://cran.r-project.org/doc/contrib/Dormann+Kuehn_AngewandteStatistik.pdf
  • *Documentation for the lme4 package: https://cran.r-project.org/web/packages/lme4/index.html

* indicates an open resource

Second semester (summer)

Environmental Statistics 

Module coordination

Prof. Dr. Carsten Dormann

Additional lecturers

Prof. Dr. Markus Weiler, Dr. Jens Lange, Dr. Kerstin Stahl

Teaching and learning methods

Lectures, practical exercises, group work

Type of examination

Project work (last days of the course), in each subject protocol of the exercises

Prerequisites

  • Basic statistical knowledge: distributions, maximum likelihood, regressions; ANOVA, GLM, PCA

Syllabus

This module builds on and extends statistical knowledge and its application:

  • Generalised Additive Models
  • Classification & Regression Trees (incl. randomForest and BRT)
  • Non-parametric statistic (resampling approaches)
  • Model selection incl. cross-validation
  • Spatial statistics (correlogram, variogram)
  • Extreme value statistics
  • Time-series analysis (autocorrelation, decomposition)

All topics will be taught in the free software R.

Learning goals and qualifications

Students will

  • extend their statistical knowledge
  • solve complex statistical tasks
  • advance the use of R

Reading

  • Crawley (2007) The R Book. Wiley.
  • *Helsel & Hirsch (1992) Statistical Methods in Water Resources. (www.epa.gov/region9/qa/pdfs/statguide.pdf)
  • Schönwiese (2006) Praktische Statistik für Meteorologen und Geowissenschaftler, 4. Aufl., Bornträger
  • *R-documentation under http://cran.r-project.org/other-docs.html, like http://cran.r-project.org/doc/contrib/Dormann+Kuehn_AngewandteStatistik.pdf

* indicates an open resource

GIS Plus 

Module coordination

Dr.-Ing. Holger Weinacker

Additional lecturers

Mirko Mälicke, Joao Paulo Pereira

Teaching and learning methods

Lectures, practical exercises, self- studies with homework

Type of examination

Exercises, homework and project

Prerequisites

Basic knowledge in GIS

Syllabus

In this module we will develop Python programs to fit and automate processing tasks within GIS. The focus of this course is NOT lying on the usage of already existing processing chains within a GIS system, but in the independent programming of individually adapted implementation concepts.

  • Introduction in the programming language Python
  • Analyses of environmental data (for example height data) using python programs, which will be developed in the course by the students themselves.
  • Automation of evaluation- and analysing processes within the GIS domain using Python

Learning goals and qualifications

Students will

  • extend their GIS knowledge
  • solve complex tasks concerning geo data processing based on Python
  • become acquainted with open GIS software/libraries as alternative to commercial products

Reading

Handouts and data will be provided.

Modelling Environmental Systems 

Module coordination

Prof. Dr. Carsten Dormann

Additional lecturers

Dr. Hans-Peter Kahle

Teaching and learning methods

Lecture with exercises, excursion

Type of examination

Graded exercises and homework; written test

Prerequisites

  • Basic statistical knowledge (B.Sc. level: distributions, likelihood)
  • Data import and simple statistical analyses in R (www.r-project.org)

Recommended for students who have attended the M.Sc.-module „Feldökologie“, or “Ecosystem Management“ or “Numerical Process Modelling”

Syllabus

The module teaches skills required for the simulation of environmental processes and applies them to specific systems:

  • Introduction to system models (processes, states, feedbacks)
  • Developing an understanding of an existing model based on the publications and manuals (e.g. forest growth, world economy, …)
  • Model parameterisation
  • Sensitivity analysis
  • Uncertainty analysis using Monte Carlo simulations
  • Introduction to the modelling of forest growth at the level of the single tree and the stand, using empirical, process-based and hybrid models
  • Introduction to the modelling of tree quality
  • Simulation of environmental and management scenarios with a forest growth model

All analyses will be taught in R as well as dedicate modelling software.

Learning goals and qualifications

Students will

  • understand the aims, uses and limitations of system models
  • aim generic and transferable technical skills on the use of system models
  • get the ability to judge the importance of experimental and observational data for the development and calibration of system models
  • get the ability to judge the usefulness and importance of environmental models for the management of natural resources, using forests as example

Reading

  • Bossel (2004) Systemzoo 2 - Klima, Ökosysteme und Ressourcen.  Books on Demand
  • Landsberg, J.J., Waring, R.H., Coops, N.C., 2003. Performance of the forest productivity model 3-PG applied to a wide range of forest types. Forest Ecology and Management 172: 199-214.
  • Nagel, J., 2012: Waldwachstumsmodell BWinPro http://www.nw-fva.de/~nagel/
  • Pretzsch, H. 2001. Modellierung des Waldwachstums. Parey, Berlin. 341 S.
  • Soetart & Herman (2009) A Practical Guide to Ecological Modelling – Using R as a Simulation Platform. Springer.
  • *R-Dokumentation unter http://cran.r-project.org/other-docs.html,
  • *Petzold, T. Konstruktion ökologischer Modelle mit R; http://hhbio.wasser.tu-dresden.de/projects/modlim/doc/modlim.pdf

* indicates an open resource