Riparian forest habitat structure by hierarchical class modelling - Donauauen

Suggested by: Barbara Riedler, Stefan Lang

Background

Riparian forests are among the most complex ecological systems and most important biodiversity hotspots worldwide, but are at the same time amongst the most threatened European ecosystems. Management of such protected areas requires detailed knowledge about the occurring habitats. Earth Observation (EO) data can thereby compliment in-situ data and support the monitoring of protected areas through their advantage of area-wide coverage, overcoming the problem of accessibility and the option of regular, cost-efficient updates. Habitat classification based on species classification, their composition and age classes were applied for a variety of vegetation types using different EO data. Such habitat classification and modelling can be the basis for further analyses such as the assessment of habitat quality, its conservation status, monitoring of biodiversity or the evaluation of conservation measurements like water restoration projects.

Short description

The aim of the MSc thesis is to automatically classify habitats of the riparian forest of the Donauauen, Austria using very high resolution (VHR) satellite imagery. Previous work in the Salzachauen showed that single tree species classification based WorldView-2 imagery is possible using an OBIA (object based image analyses) approach. Based on such a single tree detection, habitat modelling following different habitat classification schemes (Habitats directive, EUNIS) leads to area-wide coverage and harmonized trans-boundary results of habitat classification, which in turn can be the basis for the assessment of habitat quality. As every riparian forest has its own characteristics (e.g. different tree species composition, age structure, water regime,….) the transferability of existing work and rulesets has not been proven and should be implemented in this work.

.

Literature

• Strasser, T., Lang, S., Riedler, B., Pernkopf, L., Paccagnel K., 2014, Multiscale Object Feature Library for Habitat Quality Monitoring in Riparian Forests, Geoscience Remote Sens. Lett. IEEE, 559-563.
• Riedler, B., Pernkopf, L., Strasser, T., Lang, S., Smith, G., 2015, A composite indicators for assesing habitat quality of riparain forests derived from Earth observation data, Int. J. Appl. Earth Observ. Geoinf. 37, 114-123.
• Riedler, B., Lang, S., in press, A spatially explicit model of habitat quality, integrating spatio-structural indicators, Ecological Indicators, http://dx.doi.org/10.1016/j.ecolind.2017.04.027

Start

any time

Prerequisites / qualification

basic knowledge in e-cognition / OBIA
interest and knowledge in ecology / botany is helpful

Investigation of weather and bicycle counting data

Suggested by: Martin Loidl, Bernhard Zagel

Short description: Bicyclists are more exposed to temperature, precipitation and wind than public transit passengers or car drivers. Thus, we can hypothesize that weather and seasonal effects impact the number of bicyclists. In a previous study we investigated the attitude of bicyclists toward winter cycling (see http://gimobility.sbg.ac.at/paper_uploads/umfrage_winterradfahren.pdf for a report in German language). We are now interested in the spatial, temporal and spatio-temporal correlations between weather and bicycle counting data. Moreover, causalities that explain potential correlations should be investigated through surveys and/or qualitative research.
We can provide data from permanent counting stations in Salzburg in close to real-time and weather data in very high resolution. Alternatively, available data from other cities can be used as wel. The master thesis project is aimed to be integrated in the research efforts of the GI Mobility Lab.

References, suggested reading:

• BÖCKER, L., DIJST, M. & PRILLWITZ, J. 2013. Impact of Everyday Weather on Individual Daily Travel Behaviours in Perspective: A Literature Review. Transport Reviews, 33, 71-91.
• ORCORAN, J., LI, T., ROHDE, D., CHARLES-EDWARDS, E. & MATEO-BABIANO, D. 2014. Spatio-temporal patterns of a Public Bicycle Sharing Program: the effect of weather and calendar events. Journal of Transport Geography, 41, 292-305.

Start/finish by: Anytime

Prerequisites/qualifications: Profound skills in data management and processing. Statistical knowledge.

Evidence-based planning and optimization of bicycle infrastructure

Suggested by: Martin Loidl, Bernhard Zagel

Short description: The provision of adequate infrastructure is crucial for promoting active mobility, especially bicycling. Consequently, cities that aim to increase the modal share of bicyclists are planning to build and improve dedicated bicycle infrastructure. However, the positive effect of these efforts can only be gained when the infrastructure is connected and embedded into a seamless network. In contrast to this “ideal world vision”, decisions in infrastructure planning and building are often made based on pragmatic reasons and personal/subjective judgments.
Geographic information systems can help to provide an evidence base for infrastructure planning and optimization. For this the GI Mobility has already developed an indicator-based assessment model, which allows quantifying the quality of the road space in terms of bicycle safety and comfort. Based on this model a workflow should be conceptually developed and tested that consists of status-quo analysis, suggestions for corridors and network hierarchies as well as of effect estimation. The model interaction and the communication of results should preferably be implemented in a web environment.
Additionally to the model and pre-processed road data, we provide highly detailed socio-demographic data. Linking this research to the current FamoS (development of a high resolution bicycle flow model) project is possible and appreciated.

References, suggested reading:

• http://zgis229.geo.sbg.ac.at/network-assessment/index.html 
• LOIDL, M. & ZAGEL, B. 2014. Assessing bicycle safety in multiple networks with different data models. In: VOGLER, R., CAR, A., STROBL, J. & GRIESEBNER, G. (eds.) GI-Forum. Salzburg: Wichmann.

Start/finish by: Anytime.

Prerequisites/qualifications: Skills in data management and analysis as well as in scripting/programming are required. Experiences from any planning domain are advantageous.

Personalized and adaptive bicycle and/or pedestrian routing

Suggested by: Martin Loidl, Bernhard Zagel

Short description: Routing applications for public transit and cars consider road users as “machines”. Sometimes this approach is transferred to routing applications for bicyclists and pedestrians. In doing so, the personal abilities and preferences (which usually play a role for motorized road users) are largely neglected. Common routing applications optimize routes for average bicyclists and pedestrians. However, this average user doesn’t exist and the range of preferences is huge (compare a bike courier with a child on a bicycle for instance). In this master thesis research models and routing environments should be developed that account for individual preferences of users and have the ability to adapt to specific situations (e.g. heavy rain). The identification of a suitable approach (multi-criteria routing, model-based impedance manipulation, consideration of user feedback etc.) is subject to research.

References, suggested reading:

•  PRIEDHORSKY, R., PITCHFORD, D., SEN, S. & TERVEEN, L. 2012. Recommending routes in the context of bicycling: algorithms, evaluation, and the value of personalization. ACM 2012 conference on Computer Supported Cooperative Work. Seattle, Washington, USA: ACM.

Start/finish by: Anytime.

Prerequisites/qualifications: Skills in data management and analysis (especially network analysis) as well as in scripting/programming are required.

MSc: Destroyed villages monitoring using Copernicus Fire service, OSM data and VHR satellite imagery

Suggested by: Lorenz Wendt, Stefan Lang, Martin Sudmanns

Background

The aim of the EO4HumEn+ project is to develop and demonstrate innovative applications for satellite image interpretation and GIS to meet the diverse information needs of the humanitarian community. One task is to detect, to quantifiy, and to qualify the destroying of human settlements in conflict situations, as an indicator for population displacement. The current Rohingya crisis in Myanmar and Bangladesh impressively shows how the relation between settlement destroyment and people movement may look like.

Short description

This tool combines locations of villages, from OSM, global urban footprint or other sources like the Bill and Melinda Gates data from Nigeria with alarms from the Copernicus EFFIS system, which detects fires. Then, it selects cloud-free before and after images from S-2, and makes an automatic analysis for burnt areas at a better spatial resolution than EFFIS (with SemEO tools) . If there is an overlap, the system could then check for available VHR imagery from commercial providers. As test data, we have village points from Darfur via MSF and AI, from Myanmar (HRW), or the point data of villages of Nigeria could be used, although I do not know if burning down villages occurs there.

Project context

EO4HumEn+ (Extended EO-based Services for Dynamic Information needs in Humanitarian Action)

User organisation(s)

Médicins Sans Frontières (MSF), Amnesty International (AI), Human Rights Watch (HRW), and other NGOs active in the humanitarian domain

Start

any time

Prerequisites / qualification
Keen to collaborate in the GeoHumanitarian Action team at Z_GIS, passionate in using geospatial technologies for the humanitarian domain, interest in integrating EO data with other geospatial data sources, and using web-based processing tools.

MSc: Knowledge-based dwelling detection and categorisation

Suggested by: Stefan Lang, Petra Füreder, Dirk Tiede

Background

The aim of the EO4HumEn+ project is to develop and demonstrate innovative applications for satellite image interpretation and GIS to meet the diverse information needs of the humanitarian community. In particular, we develop methods to monitor the population dynamics in urban settlements during crisis situations, improve the analysis capabilities of the large amounts of freely available Earth Observation data by implementing partly automated routines, in particular for estimating affected population in crisis situations.

Short description

In nearly daily routines, we provide dwelling counts or estimates based on a combination manual interpretation and automated routines. Based on an existing dwelling/building typology with spectral and structural parameters, the thesis aims at developing a knowledge-based delineation and categorization routine in an object-based classification environment (eCognition).

Project context

EO4HumEn+ (Extended EO-based Services for Dynamic Information needs in Humanitarian Action)

User organisation(s)

Médicins Sans Frontières (MSF), and other NGOs active in the humanitarian domain

Start

any time

Prerequisites / qualification
Keen to collaborate in the GeoHumanitarian Action team at Z_GIS, passionate in using geospatial technologies forthe humanitarian domain, Advanced Remote Sensing course.

One question maps – Learning form simple spatial analyses

Suggested by: Barbara Hofer

Short description:Given the long-term vision to automate spatial analysis workflow development, we need to learn more about possible and meaningful combinations of spatial analysis operations (Hofer 2017). Within that context, the objective of this master thesis project is to develop an easy to use mobile application for ‘one question maps’. The target user group are tourists who are unfamiliar with the surroundings and would like to know how to get to a destination by public transport, where coffee shops are located etc. – i.e., tourists, who would like to answer ‘one question’ with spatial reference. The app has to support the generation of such one question maps that integrate (open source) spatial data and basic spatial analysis functionality to answer the tourists’ questions. Special attention needs to be paid to the user interface of the app and the presentation of the available analysis operations.

Besides the design and development of the app itself, a second topic is the development of a recommender system: As users work with the app, the app records the spatial analysis operations used for generating maps together with a user rating of the resulting map. Based on an evaluation of the collected background information, the app can learn from past experiences of users and recommend operations for specific input data.

References/Suggested reading:

• Andrea Ballatore , Michela Bertolotto, Personalizing maps, Communications of the ACM, v.58 n.12, December 2015  [doi>10.1145/2756546]
• Hofer, B., Papadakis, E., and S. Mäs (2017): Coupling Knowledge with GIS Operations–Benefits of Extended Operation Descriptions. ISPRS International Journal of Geo-Information: 6(2), 40.
• Vahedi, B., W. Kuhn and A. Ballatore (2016). "Question-Based Spatial Computing - A Case Study". T. Sarjakoski, M. Y. Santos and L. T. Sarjakoski. Berlin, Springer: 37-50.

Start: anytime

Prerequsites/qualification:
Interest in user interface design, application development and related topics