Indice Sezioni


Laboratorio di Geomatica - Geomatics Workbooks

[GW Home Page]    [GW Index]
[n° 1]    [n° 2]    [n° 3]    [n° 4]    [n° 5]    [n° 6]   
[n° 7]    [n° 8]    [n° 9]    [n° 10]    [n° 11]    [n° 12]   

GW 2 - 2nd Italian GRASS users meeting proceedings


copertina GW 2

Authors Title
Aime A., Ascari S., Bonfatti F., Monari P.
Abstract
ISOLA project environmental information system
Full Text
Altobelli A., Feoli E., Boglich Perasti B.
Abstract
Application of Goodall's Affinity Index in Remote Sensing Image Classification
Full Text
Altobelli A., Feoli E., Napolitano R., Cinco M.
Abstract
Remote Sensing/GIS techniques for risk assessment of Borrelia burgdorferi infection
Full Text
Biagi L., Brovelli M.A., Scuratti M.
Abstract
Enviromental noise modeling within an open source GIS
Full Text
Biagi L., Negretti M.
Abstract
A new approach to the polygons rasterization in GRASS
Full Text
Bianchin A., Foramiti S.
Abstract
Implementing a procedure to extract urban areas based on Multispectral Classification and Mathematical Morphology
Full Text
Ciolli M., Zatelli P.
Abstract
GIS learning with GRASS
Full Text
Ciolli M., Zatelli P.
Abstract
An Italian tutorial for GRASS learning
Full Text
Fabbro L.
Abstract
Amazonia Landscape Mapping and Biodiversity Estimation
Full Text
Neteler M.
Abstract
Towards a stable open source GIS: Status and future directions in GRASS development
Full Text
Neteler M.
Abstract
Volume modelling of soils using GRASS GIS 3d tools
Full Text
Rigon R., Cozzini A., Pisoni S.
Abstract
Looking for a new method of estimating solid discharges in small alpine watersheds
Full Text
Rigon R., Cozzini A., Pisoni S.
Abstract
Getting the Rescaled Width Function and the Derived WGIUH
Full Text
Sboarina C.
Abstract
Creation of climatic maps for Trentino from spread data
Full Text


The Abstract

ISOLA project environmental information system
Andrea Aime, Silvia Ascari, Flavio Bonfatti, Paola Monari
The ISOLA project, funded in the framework of the European LIFE/Environment programme and co-ordinated by the municipality of Modena, is aimed at providing methods and tools to support environmental control and planning for a medium-sized town. The expected project results are:

  • a sound methodology, expressed in form of guidelines addressed to the end users that plan to adopt a disciplined way of facing the environmental problem.
  • a software package to support the methodology and make its application directly feasible and controlled by end users not expert in GIS technology.

From the methodological point of view, ISOLA is based on three main activities:

  • Eco-balance. Assessment of risks in the urban environment and its components, and measurement of the pressure exerted by human activities and settlements.
  • Eco-plan. Integration of environmental factors in urban planning according to proper methods, criteria and procedures.
  • Eco-management. Support to decision-making by means of simulation tools that allow evaluating the impact of alternative scenarios.

Every methodological activity requires some kind of spatial analysis, these activities need to be carried out as a result of the collaboration several experts in different environmental issues, and are carried out on a periodical basis: this means that the software system, named SIA (Sistema Informativo Ambientale, that is, Environmental Information System) has to put a strong focus on spatial data analysis and representation, information sharing and analysis automation. Since ISOLA project is funded by a Europeand programme results have to be useful for every other medium sized municipality, thus exportability of software, in terms of cost, technology and automated activities is a major concern.
In this paper we outline the main problem that SIA has to face and the solutions found in order to effectively face competing and apparently incompatible requirements. Section 2 provides a more in depth analysis on the issues that a spatial analysis system has to face in order to be versatile, user friendly and effective. Section 3 describes the process model, a graphical tools providing a good compromise between automation and user friendliness. Section 4 describes the client server architecture that enables user to share both data and processes in a cost effective way by leveraging as much open source software as possible, and in particular how GRASS (3) is being used to provide the most important foundation to environmental data analysis, while Section 5 draws conclusive remarks and gives some hints about future improvements.
Top the List

Application of Goodall's Affinity Index in Remote Sensing Image Classification
A. Altobelli, E. Feoli, B. Boglich Perasti
The affinity index of Goodall is introduced to integrate remotely sensed data with multisource data (generally called ancillary data) to improve land cover classification. The classification through Goodall's affinity index is performed using "r-affinity", a computer program offered in the GRASS-GIS environment which can deal with quantitative maps (such as image or DTM data), qualitative data (such as landcover maps), ranked data (such as a slope class map). Furthermore this program doesn't require normal distribution, like others popular classifiers used within a GIS context. In this paper "r-affinity" is applied to a set of data from a plateau-area surrounding Trieste and the results are compared with those obtained from maximum likelihood (carried out using GRASS). The highest overall accuracy is achieved by Goodall's affinity index; in particular, while the classified image obtained by the other method look exceedingly "smoothed" (over-smoothed), the one obtained by "r-affinity" is much more reliable (meaning "near to real land cover situation") and therefore more suitable and efficient for ecological researches.
Top the List

Remote Sensing/GIS techniques for risk assessment of Borrelia burgdorferi infection
A. Altobelli, E. Feoli, R. Napolitano, M. Cinco
The associations between the presence of infection of Ixodes ricinus with Borrelia burgdorferi and environmental variables can be quantified using remote sensing data processed in a Geographic Information System (GIS).
The risk of Borrelia burgdorferi infection was evaluated in a small recreational area located on the karst plateau of Trieste, representative of the whole karst. Ixodes ricinus ticks, collected in different ecotypes monthly throughout the year (September 1997 - October 1998), were tested for the presence of Borrelia burgdorferi.
The highest presence of ticks (60%) and infected ticks (38%) was found in dolines, a typical depressions of the karsic landscape characterised by high humidity and deciduous woodland.
Concerning the faunistic analysis, particular importance was given to small rodents (reservoir of infection) and to the presence of roe deers (Capreolus capreolus), which are very common ungulate in the karst territory. Since the distribution of small rodents did not vary significantly from one ecotype to another and roe deers are ubiquitous in the study area (20-30/100 ha.) the research focused on vegetation, climatic and geomorphologic considerations.
A Landsat 5 Thematic Mapper image (31/07/97) was used to extract the moisture index (TM4/TM1), the surface temperature and vegetation maps and to identify the most favourable habitats for Ixodes ricinus.
Using the cartographic data on urban areas and escursionistic tracks, it is possible to identify potential infection risk areas.
Top the List

Enviromental noise modeling within an open source GIS
Ludovico Biagi, Maria A. Brovelli, Marco Scuratti
The acoustic pollution caused by vehicular traffic in urban environment has such a large effect that it can't be further ignored. In order to protect the Public Health, every Italian Municipality, following the local D.P.C.M. 1/3/1991 law, must split its territories into homogeneous areas depending on the largest acoustic noise level accepted ( acoustic zoning).
This work, developed in the frame of the I.S.O.L.A. project ( Information System for the Orientation of Local Action ), is concerned with the Local Actions that have to be undertaken to reach a sustainable development in an urban environment. Our aim is to provide support to the acoustic suffering analysis, defined as the comparison between measured noise values and largest values allowed by the laws in force.
The steps of the work are:

  1. Determination of a regression model, using a robust statistics, based on the punctual noise measurements and some traffic parameters like mean speed, hourly mean flow classified by vehicle typology. The model, taken from literature, is described by a non-linear equation to be linearized and iteratively solved. The approach is similar to a least squares procedure, but in this case we use robust estimators, less sensible to the presence of outliers. The relationship between the observables was tested by variance analysis. In such a way we were able to identify significant different speed classes and related set of parameters. This step ends by classifying each road segment by the level of a noise.
  2. Implementation of a GRASS command, allowing the noise computation in a direction perpendicular to the road axis. Here the buildings are taken into account as obstacles, while their reflection effects are ignored. The program can be used to estimate the noise field from the noise values associated to each road: if we use observed values as input, we obtain an estimate of the actual pollution; otherwise we can obtain different scenarios introducing simulated noise values as input.

Top the List

A new approach to the polygons rasterization in GRASS
Ludovico Biagi, Marco Negretti
Within the GIS framework the rasterization of polygons vectorial themes is needed when a raster approach has to be applied in the theme analysis: for instance in the overlay between some normative zones (i. e. a polygon vectorial theme) and one or more environmental phenomena (typically represented by a raster model).
It is well known that, formally, an univocal algorithm doesn't exist; the most popular approaches are the center, the area prevalence and the importance methods. Up to now in GRASS software the first approach was the only one available.
Because of the numerous environmental applications that require one of the other two methods, we have studied and implemented an "approximate" version of both. This is realized by using the center method on a working raster, characterized by an higher resolution than the final one; thus the value of every final cell is computed by evaluating the working cells contained in it. The user can choose the method and the computation approximation, e. g. the ratio between the final and the working raster resolutions; for the importance method an "importance scale" must be specified too. In the paper the approach and the results obtained both on example and on real data analyses are discussed.
Top the List

Implementing a procedure to extract urban areas based on multispectral Classification and Mathematical Morphology
Alberta Bianchin, Stefano Foramiti
The operation of extracting urban areas has been completely developed with GRASS running under Linux. The whole procedure has been implemented through script (bash e awk) containing iterative and conditional instructions.
The multispectral image Landsat5 TM dated 14th July 1994 has been used to extract urban areas of Treviso district. The CTRN images of Treviso have been rectified and placed on the Landsat location, therefore used as training areas of two classes (built, non-built) in the training stage for the supervised classification of the Landsat scene. The panchromatic SPOT-PAN image has been processed through mathematical morphology operations in order to extract urban areas. The result of this procedure has been combined with the Landsat classification in order to improve the accuracy of the final result. (Bianchin - Pesaresi, 1992).
The data preparation stage includes the following operations: selecting the study area and the suitable vector layers of the CTRN; conversion to raster of the selected layers, preparing the maps for training and test of the supervised classification of Landsat TM image, georeferencing and rectifying the CTRN layers and Landsat and Spot images.
The implementation of the mathematical morphology procedure to extract built areas has been developed by the following operations: the creation in GRASS of the location and mapset containing the Landsat images, the SPOT-PAN image and the vector and raster georeferenced maps of the CTRN; the operations of georeferencing and rectifying (i.points, i.rectify2), allowing to superimpose the CTRN grid to the Landsat grid (avoiding to alter the radiance and the co-occurrence of the adjacent pixels, effects usually produced by the resampling operation in traditional georeferencing stage); the supervised classification (i.maxlik) of Landsat TM images to extract urban areas, using the CTRN in two ways, as training map (i.gensig) and as test (r.kappa); the geometric correction allowing to superimpose the SPOT and the Landsat TM grids to the CTRN one; the operation of morphological gradient applied to the SPOT PAN image (r.mapcalc); the selective resampling with kernel 3x3 of the gradient image with a low pass filter "erosion" (r.neighbors) and the binarisation of the image based on a suitable threshold interactively determined (r.mapcalc); the intersection of the Landsat image classified (built areas) with the binarised SPOT image, in order to obtain a more accurate classification.
Top the List

GIS learning with GRASS
Marco Ciolli, Paolo Zatelli
Since 1996 GRASS is used in GIS classes at the Dipartimento di Ingegneria Civile ed Ambientale of the Università di Trento. This paper describes the potential of this GIS, with freely available code, in Geographical Information Systems education, both for university courses and professional training.
The use of GRASS offers several advantages: the software is accessible without licensing worries, the system can be freely distributed to students, the source code is available for the analysis of the algorithms and a large free collection of documentation exists on the Internet.
Educational experiences gathered in several years of teaching using multimedia devices have been applied to specific courses on GIS and in land planning and management, where GIS is used as a tool.
Different approaches in GIS training are presented: traditional lessons, tutorials and full GRASS applications.
Top the List

An Italian tutorial for GRASS
Marco Ciolli, Paolo Zatelli
A new tutorial has been developed using the experience of several courses. It is written in Italian language to help new Italian users approaching GRASS for the fist time without the language barrier. The tutorial has been written using Hypertext Markup Language (HTML) to follow the reasoning paths and logical jumps of operative practice, overcoming the stillness of traditional manuals. The aim of the tutorial is to show and describe gradually and in a friendly way the main functions of GRASS, so that GRASS learning can be fast and new users can be productive in a short time. A first working version is described, while future improvements, also with users' feedback, are planned.
Top the List

Towards a stable open source GIS: Status and future directions in GRASS development
Markus Neteler
The ongoing GRASS GIS project is following the open source development model with fully accessible source code which is released under GNU General Public License and regular releases. The basic development is undertaken by the "GRASS Development Team", special contributions are delivered by institutions or individual developers. If such contributions are interesting for a wider audience, they are integrated into the main GRASS system. In 2000 and first half of 2001 the development has been focusing on code stabilization and portability - GRASS shall be compliant to rather any UNIX derivate and MacOS X (Darwin) and, utilizing the CygWin tools, even on MS-Windows platforms. Important new features of GRASS 5.0.0 are introduced in this article. New ideas and changes will be introduced in the upcoming GRASS 5.1.0, the author briefly outlines basic ideas of proposed improvements. Due to recent changes the paper reflects latest development up to July 2001.
Top the List

Amazonia Landscape Mapping and Biodiversity Estimation
Luigi Fabbro
This project aims to set-up and test a methodology for landscape mapping of Amazonia to assess the biological diversity by applying the knowledge and know-how of the local communities and involving them in Remote Sensing and Geographic Information System technology. The project will provide ground truth for the remote sensing of Amazonia and contribute to the baseline data sets of pristine biotopes against which "trouble spots" can be monitored.
The concrete output of this project will be a GIS mapping the landscape of the Lower Rio Jauaperi area and in particular to map the presence of selected taxa belonging to the local community taxonomy. These maps may be used to determine biodiversity indexes like the alpha, beta, gamma and other measures of biodiversity so that comparable data is produced (Fabbro, 2000).
Top the List

Volume modelling of soils using GRASS GIS 3d tools
Markus Neteler
The analyses of spatial variability of soils have been of interest to soil scientists and geographers for quite some time. Information on soil properties are usually available from a limited number of point measurements and spatial estimates are prepared in two dimensions (e.g. by interpolation or other technique). However, soil is essentially a 3D object with varying properties in all spatial dimensions.
This study focuses on 3D capabitities of GRASS GIS providing new 3D tools to manipulate, analyse and model 3D landscape phenomena. For example, the multivariate interpolation method - regularized spline with tension (RST) has a capability to interpolate and analyse geometric properties of selected soil properties in three-dimensional space. We investigated the possibility of modelling dynamic processes occuring in soil using simple 3D map algebra algorithms. An inherent part of scientific investigation and analysis is a visualization. New GRASS visualization tools exploit 3D OpenGL graphics capabilities, coupling to external visualization software (Vis5D) allows animated views to time-dependant processes in soil volumes.
An increasing number of available 3D environmental data requires a complex GIS solution for manipulation, analysis and modelling. Using G3D library, new tools for modelling and visualization, GRASS has proved that it fulfils these requirements. Hence, GRASS GIS belongs among the first GIS's with full 3D analysis capabilities in the world.
Top the List

Looking for a new method of estimating solid discharges in small alpine watersheds
Riccardo Rigon, Andrea Cozzini, Silvano Pisoni
A quantitative method for the gross estimation of landslide areas, volumes and solid discharge in extreme events is proposed. The method is based on the use of digital elevation models (DEM) in a GIS environment. First the SHALSTAB model has been used (Dietrich, 1994) to evaluate the area subject to landsliding according to specified hydrological conditions, then we have coupled SHALSTAB with soil depth model to get landslides volumes. However, landslides from hill-slopes are not the only source of solid discharges in Alpine watersheds. Major damage is in fact produced by the failure of sediment accumulated in unchannelized valleys and channel beds. These sources produce, after geo-mechanical instability triggered by subsurface water flow, solid movements called debris flows which are a mixture of variable concentration of sediments and water. We propose to assume that, at least for events of medium to large return time in small watersheds, the solid discharge total volume can be derived evaluating liquid discharges with a new method. Used methodologies are partially implemented in GRASS and partially use new codes which are being ported to GRASS.
Top the List

Getting the Rescaled Width Function and the Derived WGIUH
Riccardo Rigon, Andrea Cozzini, Silvano Pisoni
It is shown how rescaled width functions (Rinaldo et al., 1995; D'Odorico et al., 1996) are built by means of Grass and some other custom programs. From them the width function based instantaneous unit hydrograph, WGIUH, is obtained. In the end, it is discussed the possible generalization of the procedure to include spatial soil moisture variability.
Top the List

Creation of climatic maps for Trentino from spread data
Chiara Sboarina
The aim of this study is to create climatic maps for one particular region, the Trentino in the north-east of Italy. The availability of a great number of measurement stations all over the region and the use of a geographical information system are the characteristics of this work. Interpolating values of climate variables from points to large areas is important in a variety of disciplines, therefore it is fundamental to minimize the extent of interpolation errors by using a suitable interpolation method. We compared the performance of three interpolation methods: Inverse Distance Weighting (IDW), Regularized Spline with Tension (RST) and Ordinary Kriging. The first two procedures are integrated in GIS GRASS while for Kriging it has been taken advantage of an external program named Gstat. Before interpolating it was found that there are no relationships between precipitations and elevation or coordinates of measure stations, while for temperature was found the known relationship with elevation. Annual average rain data were interpolated using the three different methods and the results were compared on the base of statistics and cross validation. Monthly average temperature data for all year months were interpolated after being taken back to see level. The choice of the best interpolation method for temperature was made for January and August data because they are the coldest and the warmest month of the year. The final result is the map of annual precipitation and twelve maps, one for every month, of temperature for all Trentino.
Top the List

info pdf readerSoftware per leggere i documenti
in formato pdf



Geomatics Workbooks n 2

Editorial Board: Maria A. Brovelli, Ludovico Biagi, Marco Negretti

Editor: Maria A. Brovelli





Laboratorio di Geomatica - via Valleggio, 11 - 22100 Como