Refolana summer school use of open source tools for spatial ecological modelling

CALL FOR APPLICATIONS

• 
  based on GRASS, Qgis, AWK, GNUPLOT, GDAL/OGR, under Linux Environment
  

Deadline for applications: May 1st, 2010

FEES: 300 €

CREDITS: 10 ECTS

PROGRAMME

This 2-weeks REFOLANA course aims at demonstrating different free and open sources tools to carry out spatial ecological modelling and spatial analysis in ecology.

The course comprises a set of short lectures that present modelling concepts and procedures, as well as practical sessions for participants where participants can use the software and scripting routines to achieve large-scale data analysis and stand­alone.

The first week is dedicated to lectures and exercises of the open source software. During the hands-on software practical sessions, participants will actively interact with different tools and carry out spatial data analysis by manipulating data from different case studies. Different exercises will be proposed and participants will have the opportunity to improve their computational skills in both testing ecological theory and solving practical questions while producing maps and summary statistics. Particular emphasis will be placed on forest ecosystems applications.

During the second week participants will practice their acquired knowledge on a short personal project. The participants will be supervised in their work and will have the opportunity to analyse their own dataset and solve their own project. This short project

will allow them to briefly go through the different phases of spatial ecological applications: conceptualization, data preparation, model fitting, model evaluation, spatial prediction, assessment of model applicability, application of the model results and summary statistics.

All lectures are in English

Coordination

Dr. Annemarie Bastrup­Birk,

University of Copenhagen,

Forest & Landscape Denmark,

Hørsholm Kongevej 11 DK­2970 Hørsholm

Tlf. +45 3533 1897 ab@life.ku.dk

Dr. Giuseppe Amatulli (giuseppe.amatulli@gmail.com)

Dr. Stefano Casalegno (stefano@casalegno.net)

Dr. Annemarie Bastrup­Birk (ab@life.ku.dk)

Dr. Patrik Nyed Karlsson (pnka@life.ku.dk)

Lucia Seebach (lucia.seebach@jrc.ec.europa.eu)

At the end of the course participants should acquire a basic knowledge of a selection of open source tools that are available in the context of spatial ecological modelling and GIS/spatial analysis. After this training the students should be able to progress independently and learn how to process data using open source software. The course gives a useful background for data analysis in the context of forest, biodiversity, conservation, landscape planning and other related topics such as precision farming, geo­statistic, socio­economic issues, remote sensing and non­spatial modelling.

Linux shell scripting, GRASS and Qgis geographic information systems, R language and environment for statistical computing and graphics, AWK programming language for processing text­ based data, gnuplot program for two­ and three­dimensional plots of functions and data, geo-tools library for the manipulation of geospatial data (Gdal and OGR translator library for raster and vector geospatial data).

This training is addressed to a diverse population of students at master's or doctoral level with a common interest on spatial data analysis and ecological modelling.

Students should have a basic knowledge of GIS to be able to follow the training and better improve their computational skills.

Students are expected to provide an idea for a one week project as well as needed data

to be processed in order to fulfil the project task. If needed a project proposal with its related data

can be provided in agreement with students and with the responsible of the Department of Forest & Landscape.

In order to obtain the 10 ECTS (European Credit Transfer System) the participants should present the project and the results at the end of the course - or submit a paper (min. 10 pages) on the results at the latest 1 month after the end of the course.

Appropriate exercises will be provided allowing students to practice the use of tools and methods in spatial data handling. The exercises and examples are applied within a large variety of topics: forestry, landscape planning, predictive modelling, mapping, nature conservation and spatially related fields of study. In agreement we can adjust key examples and focus exercises on other special needs.

1. 
   Knowing each other: trainers and student's background
   
2. 
   Identifying student's needs
   
3. 
   Course objectives and schedule
   

1. 
   Linux environment
   
2. 
   Why and what to use
   
3. 
   Introducing the Ubuntu virtual machine
   

1. 
   The basic modelling concepts and procedures in ecology
   
2. 
   Introduction to Spatial ecological modelling
   
3. 
   Models examples: non­parametric algorithms, ensemble classifiers, machine
   
4. 
   learning techniques
   
5. 
   Forest habitat suitability modelling
   
6. 
   Forest fire danger forecast model
   

1. 
   Why to use Unix/Linux command line
   
2. 
   The basic commands
   
3. 
   Command syntax
   
4. 
   File management
   
5. 
   Read and explore a text file
   
6. 
   Meta­characters and their use
   
7. 
   Concatenate process
   
8. 
   The use of variable
   
9. 
   The ‘for’ loop
   
10. 
    Case study and exercise
    

1. 
   Why to use AWK command line
   
2. 
   The basic commands Command syntax
   
3. 
   Built in variables
   
4. 
   Import variables
   
5. 
   String functions
   
6. 
   Numerical functions
   
7. 
   Query function
   
8. 
   Manipulate large files before importing in R
   
9. 
   Case study and exercise
   

1. 
   Why to use GNUPLOT command line
   
2. 
   Commands syntax
   
3. 
   Plotting data
   
4. 
   Use AWK language inside GNUPLOT
   
5. 
   Case study and exercise
   

1. 
   Introduction to GDAL/OGR
   
2. 
   Command syntax
   
3. 
   Raster data manipulation
   
4. 
   Built up specific geo­tools
   
5. 
   Case study and exercise
   

1. 
   Introduction to R environment
   
2. 
   R structure, libraries, scripting and getting help
   
3. 
   Command syntax, R objects
   
4. 
   Basic commands (input, output, data creation)
   
5. 
   Data manipulation
   
6. 
   Plotting data and graphical parameters
   
7. 
   Programming (functions, if and if else condition, for loop, while, system variable)
   
8. 
   Spatial and ecological modelling libraries Basic statistic
   
9. 
   A linear model and step­wise regression
   
10. 
    More complex algorithms (Classification and Regression Tree / Random Forest)
    
11. 
    Importing geo­data
    
12. 
    Model prediction
    
13. 
    Exporting the data
    
14. 
    Case study and exercise
    

1. 
   Introduction to Quantum gis
   
2. 
   Import and export data
   
3. 
   Data visualization and map creation
   
4. 
   Qgis as Gui to GRASS and as a learning tool
   

1. 
   Introduction to GRASS
   
2. 
   Data structure in GRASS
   
3. 
   Command syntax and general commands of data handling
   
4. 
   Raster and vector data import, export, display and conversion
   
5. 
   Raster map calculator
   
6. 
   Vector manipulation and processing
   
7. 
   GRASS / Qgis interface
   

8. 
   Scripting in GRASS: combine BASH commands with GRASS commands
   
9. 
   Environment variables and their importance for shell scripting
   
10. 
    Case study and exercise
    

1. 
   Projecting future burnt area in the EU countries under IPCC SRES A2/B2
   
2. 
   Modelling the Distribution of the Natural Forest Formations in Europe under present and future climates
   

1. 
   Students presenting their case study dataset and related questions:
   

2. 
   Open discussion on students' interests and new tools to approach case study
   
3. 
   Identification and statement of students' needs, method of data processing and
   

4. 
   Data preparation, statistical data analysis, modelling and mapping, synthesis and
   

5. 
   Students presenting their project and results
   

The training is a hands-on oriented approach. Classes will be carried at The University of Copenhagen, Forest & Landscape, Frederiksberg in rooms with a video projector and access to computers connected to the internet for each student.

Before the course starts the instructors will install a Linux Ubuntu operating system ­ virtual machine (http://www.gisvm.com/) and all required software on every computer.

A wiki web­site infrastructure will serve as a tool for data exchange, download scripting procedures, download case study and exercises, uploading exercise results and group­work.

Each student will receive an account on the wiki and will be able to use it as working tool.

The web hosting is provided by the instructors.

Instructors: G. Amatulli, S. Casalegno, A. Bastrup-Birk, L. Seebach (P: students' project)

climate data. Modeling forest species distribution, current, potential and future trends using WorldClim dataset. He is dealing with high resolution data in the context of complex andmodern modeling techniques by stand­alone implementation process.

statistical analysis for fuel model mapping and fire occurrence assessment. Before joining the JRC he was working at the Universidad de Zaragoza ­ Spain ­ Dep. de Geografia as a Post­Doc researcher scientist.

plant bird interactions (Univ. Paris Sud and Univ. of Bristol, Kibale Forest Nat. Park Uganda, 1996); Mountain Forest Ecotones and vegetation belt mapping, (CEFE ­ CNRS, France 1997, CIBNOR La Paz ­ Mexico 1998, INA P­G France 1999­2001); Biofuels, forestation potential and forest yield estimations (INRA Versailles ­ MAICh Crete 2002);

Integrated studies for the assessment of forest condition and climate change impact on

European forests (JRC of the European Commission 2004­2008). Recently, he has been working on suitability modelling of European forests, Urban and peri­urban forests, and alpine ecosystems. He is involved in the activities of the International Union for Forest Research Organizations and participating in the European expert panel on forest growth of the ICP­Forest task force (United Nations Economic Commission for Europe).

PhD in Applied Forest Ecology (GIS ­ remote sensing) at INA P­G, France (2001). MSc Ecology “DEA d’Ecologie” Eco­physiology and Ecosystem functioning at INA P­G Univ. Paris­Sud XI and Univ. Paris VI (1997). BSc Ecology. “Maitrise d’Ecologie”, Univ. of Paris­Sud XI (1996). University degree in Natural Sciences “Licence ès Sciences Naturelles”, University of Paris­Sud XI (1995).

B.Sc in Geography and Geoinformatics at the Institute of Physical Geography, University of Copenhagen. MSc. in Agronomy at the Institute of Agrohydrology and Bioclimatology, University of Agriculture. PhD. in agronomy at the Institute of Agrohydrology and Bioclimatology, University of Agriculture, Copenhagen.

DEADLINE May 1^st, 2010

COURSE FEES 300 €

The registration fees includes course fees, coffee/ tea, conference material

Participants will be notified for acceptance. Further information on venue and accommodation will be sent after acceptance

Payment is requested by either

Bank transfer to Danske Bank – mark your name and use of.
Account No: 3306-4490058141

IBAN NR. DK6330004490058141