Pierre Racine

Research Professional
Geographic Information Systems & Analysis/Information System Architecture/Programming

Pavillon Abitibi-Price, Université Laval
2405, rue de la Terrasse
Bureau 2171
Québec, QC, Canada, G1V 0A6

(418) 656-3731

This page: http://bit.ly/ppx8VZ 
My blog: Geospatial Elucubrations 

ACADEMIC BACKGROUND

• M.Sc. in geography, unfinished (Université Laval, 2006)
• B.Sc. major in mathematic and computer sciences and minor in journalism, 1994 and 2004 (UQAC and University Laval)
• Many courses in theoretical computer science (Université Montpellier II, France, 1995)

ABOUT ME

After two years at the B.Sc. in mathematics and computer science in Chicoutimi and a year in Montpellier (France) studying theoretical computer science, I worked as a programmer analyst for a few artists (see Meduse ) and in the software industry for Bentley Systems . My first work for Bentley was to port the vectorization software Descartes to UNIX workstation. I then became technical lead of the overhaul of their main geoserver product: Bentley Publisher 3.0. It was mainly used to publish drawings (DGN) and geographic projects on the web.

After some wanderings between Paris and Dar es Salaam and others in journalism in Quebec City, I came back to science by enrolling in the Master’s degree in Geography at University Laval. My project, entitled "Modeling of biographical itineraries in a spatio-temporal GIS", was to find ways to store routes of historical figures in a spatial database (PostGIS / PostgreSQL) and to map them online. The main challenges were to develop a data model to capture the spatial and temporal information related to the event happening to characters along their life path, to develop a method to capture imprecise moments and locations using fuzzy logic and to map these networks and travel at different levels of accuracy in an online map (with MapServer and Google Earth among other packages).

RESEARCH INTERESTS

• Transparent Raster / Vector Spatial Analysis - Most GIS packages provide two sets of tools for spatial analysis: one for doing analysis in raster mode and another one for doing analysis in vector mode. However, operations that analysts want to do on these two types of data structure are essentially identical. The PostGIS Raster project (see description below) aims to provide an elegant solution to this problem by giving (desktop or web) application developers a single SQL manipulation and analysis API working transparently on raster and vector data.
• Wiki Collaborative Information Systems - Content Management Systems (CMS) are very useful to build websites but structured conceptual and terminological differences between CMS (pages, templates, extensions, articles, etc ...) make the transition from one to another very difficult. On the other side, even if wikis share a similar set of concepts that make them easy to learn and use, they do not offers easy services to build structured content. A CMS-Wiki hybrid, XML based publishing system, offering the best of both worlds would almost be the perfect system: A very simple and flexible concept for structuring content integrating the most fundamental notions of data management and a very quick and simple web 2.0 graphical edition interface (using context menu) inspired by wiki systems.
• Spatially Explicit Multi-Agent Simulation Systems – Still, no multi-agent simulation system based on a GIS is flexible enough to be used by different disciplines (ecology, biology, epidemiology, management, forestry) and easily expandable so that a community of user can develop libraries of reusable agents very much like what the R community has accomplished with its vast library of packages and functions. Some simulation packages are getting near to this ideal like Repast or NetLogo but the first requires very good programming skills (C or Java) and the second is quite limited in terms of model expressivity.
• Modeling as a Scientific Process - Modeling is increasingly at the heart of the scientific discourse. We believe it is actually at the heart of all scientific activity, all disciplines combined. It is time to beter define what is modeling, what are the basic types of models and to model (!!!) the scientific process as a modeling activity. This could help us to standardize and automate the process of scientific inquiry in a rigorous manner.

TEACHING

I sometimes teach to members of the CFR.

• Intensive introduction to ArcGIS
• Introduction to geographical information system with QGIS
• How to make your web page on the CFR website

CURRENT PROJECTS

• PostGIS Raster  - PostGIS Raster’s goal is to provide support for storage, manipulation and analysis of images data (rasters) to the PostgreSQL/PostGIS relational spatial database. This is a major Open Source project that has attracted the collaboration of many public and private organizations from many countries.
  
  PostGIS Raster implements the type RASTER in the most integrated and transparent way possible with the SQL API for PostGIS already available to support vector features (the GEOMETRY type). It offers overlay functions (such as ST_Intersects or ST_Within) working transparently without the user having to worry if the layers involved are in vector format or in raster format. It also includes a complete set of functions for processing and analysis of rasters.
  
  PostGIS Raster introduces thereby a new type of geospatial object, the "raster objects", which, as a polygon in a vector layer, have their own geo-reference and their own individuality (or set of properties or attributes). These new objects facilitates the modeling of surface objects having variable values (such as fires, lakes or surface deposits) generally poorly modeled by a single raster or by polygons with crisp edges. .
  
  PostGIS Raster enables researchers in ecology and environment sciences to query high-resolution rasters on extents as huge as Canada. For more details see the official website of PostGIS Raster. .
  
  PostGIS Raster was presented:
  • at the Colloque Géomatique 2009 in Montréal in october 2009 ,
  • at FOSS4G in Barcelona, Spain, in september 2010 ,
  • at the Centre de recherche en géomatique à Université Laval, Quebec City, in october 2010 ,
  • at the Centre d’information topographique, Sherbrooke, in january 2011 ,
  • at FOSS4G in Denver, Colorado, in september 2011 ,
  • at one Session PostgreSQL dedicated to PostGIS in Paris, France, in june 2011 ,
  • at the Colloque Géomatique 2011 in Montréal in octobre 2011 ,
  • at the Global Geospatial Conference in Quebec City in may 2012 .

• A Common Attribute Schema (CAS) for the Canadian Forest Resources Inventories - In Canada, each province produces, sometimes by itself, sometimes through the logging companies, its own forest inventory. These numerous forest inventories (about fifteen) varies in structures, sizes and qualities from one province to another. The objective of this project, supervised by Steve Cumming and on which we have been working from 2008 to 2012, was to gather all these inventories and convert them into a single standardized inventory.
  
  This unique inventory, covering all the Canadian boreal forest, will be much more accurate than the National Forest Inventory of Canada . It was loaded into a spatial database (PostGIS) and intersected with numerous ecological dataset used by our students in master's and doctorate degree. I supervised the work of Mélina Houle who wrote the conversion scripts (about 25) and the SQL intersection queries. Etienne Bellemare Racine and Bénédicte Kenmei Djiki worked on the standardization scripts in Perl.
  
  We have presented the project at the ESRI Regional User Conference in Quebec City in April 2011. See presentation entitled Automating the conversion of forest inventories Canadians with ArcGIS and Python .
• Canadian Data Foundry - This project aims to build a web GIS for querying different high resolution ecological layers covering the whole of Canada. Users upload their vector layers and build their spatial queries that are executed by a supercomputer. Two key elements of this project are PostGIS Raster and CAS.
• Boreal Data Foundry - The Boreal Data Foundry is a catalog of all datasets used in projects in which Steve Cumming is collaborating (BAM , BEACONS , BEEST , BATS).
• Training on the principles of programming - I am (slowly but surely) still trying to develop a three-day course on the general principles of programming. The purpose of this course is not to teach programming in a specific language but to teach programming in a very general way.

SOME PAST PROJECTS

• www.riffeac.org  for Damase Khasa - This project involved building a collaborative website for the Réseau des institutions de formation forestière et environnementale d'Afrique Centrale (RIFFEAC). Institutions members of RIFFEAC are scattered throughout Central Africa. It is therefore difficult for a single webmaster, located in one of these institutions, to produce up to date information on courses available. The new site, based on wiki technology (the same as the one used to build the CEF site), should allow webmasters from each different institution, who are well aware of each particular course, to update the common web site themselves.
• Random Transect for RebeccaTittler - Random Transect is a small extension to ArcMap (built on the same principle as Hawth's Analysis Tools) used to generate random transects within a specific region (shapefile). Rebecca was doing a postdoct with Christian Messier.
• Crown Delineator for Alain Paquette - The purpose of this project was to automate the analysis of the area occupied by a treetop from images of the summit and at different levels of resolutions. Everything was done in a Python script automating ArcGIS geoprocessing tools. Alain Paquette was doing a postdoc supervised by Christian Messier and Sylvain Delagrange.
• pyFGE  with Axel Wellpott - Adaptation and online publishing of an application allowing evaluation of the risk of windthrow. Axel’s application was only working in Handler Python mode (mod_python) under Apache. We adapted it in order to host it on our IIS server that does not support this mode. Alex was a student with Jean-Claude Ruel.
• TOPEX  - TOPEX is the abbreviation of TOPographical EXposure. It’s coefficient representing exposure to wind. TOPEX is a ArcView 3.2 project originally created by the Windthrow Research Group  led by Dr. Stephen J. Mitchell at the University of British Columbia. My work on TOPEX consisted to simplify the code and rewrite the documentation for Jean-Claude Ruel.
• The CFR website  - One of my first task at the CFR, was to build a new website for the center. We wanted all people working at the different poles of the CFR (Montreal, Quebec, Abitibi, Sherbrooke) to be able to publish news and events on the site. I thought it would also be interesting to allow every CFR member, not only regular members, to have and be able to modify their own webpage. The principal asset of an organization consist in its human resources and it is important to emphasize them, but there are more than 400 members at the CFR! It is impossible (or at least it is not very realistic) to create and maintain a web page for each of them. That's why I builded the CFR website as a wiki (PmWiki ), with some pages open in write mode to the CFR coordinators (Luc Lauzon, Melanie Desrochers, Daniel Lesieur, myself and formerly Sophie Brugerolle) and the opportunity for each member to create its own web page by himself. This principle ensures that the information contained in the site is updated by the person who best knows this information. The result is a site more up-to-date, dynamic, benefiting from the creativity of all and offering great exposure to the CFR because of the diversity of its content. Up to now, over 100 CFR members have their own web page.
  
  My role as webmaster has led me to work on some big projects for the site:
  • The online list of equipment managed by members of the CFR - Publishing, with Juan Posada, of the list of equipment owned and managed by CFR members. These pages are designed to stimulate sharing of equipment between the CFR members and thereby maximize their use. Each device has its own wiki page. This allows managers of these facilities to post and update information relevant to their use.
  • Fox Include - Fox Include is an extension to Fox, itself an extension to PmWiki. It allows storing tabular information in CSV format into a wiki page and displaying them in a page formatted with a template. This allows us separate the content from its presentation in PmWiki, a fundamental principle in information management and publication. Basically, a page containing data in CSV acts as a database which information can be displayed in one or many other page, the way we want. Fox Include allows filtering of rows and columns and sorting the data before display, very much like SQL does with a data table. I used Fox Include for several features in the website of the CFR: the pictures appearing randomly on the first page, the members profiles in the lower left corner, the member pages. A similar extension is used to store and display all the 3000 publications of our researchers.

SOME PERSONAL PROJECTS COMPLETED BEFORE MY APPOINTMENT AT THE CFR

• Representation of Biographical Trajectories in a Spatio-Temporal GIS – Master’s Degree Project - Travel books, and more generally life stories of historical figures, are of primary importance in the literature, history, ethnology and geography research. These trajectories have spatial and temporal dimensions of considerable interest from a geographic point of view. If these disciplines did not neglect the interpretative side of their studies, they paid little attention in renewing their analysis and representation using quantitative methods and new tools from information science disciplines.
  
  This research aimed to define a framework (or model) of analysis and representation of life trajectories in a temporal GIS. The implementation of this model, however, was confronted to the limits of existing information systems. These are indeed devoid of tools to capture and manipulate nominal, temporal and spatial imprecise information abounding in life and travel accounts. Mapping systems also have the drawback of only simulate, by means of several layers, multiple representation of complex paths. An adequate system must allow the representation, on the fly, of trajectories at multiple levels of detail directly from the information entered.
  
  We therefore implemented a relational model for describing life trajectories in a spatial database management system (DBMS) and we developed an extension to the SQL language to enter, manipulate and search imprecise temporal information contained in this model. This experience has allowed us to define a method for a similar development for the imprecise spatial side of the information contained in the trajectories. We also implemented an algorithm, in SQL, to generalize, on the fly, allowing the mapping of every events of a biographical trajectory whatever their levels of detail.
  
  The first data used were from the first book of Samuel De Champlain recounting his trip to the West Indies.
  
  -Master’s Main Talk  (June 2006).
  -Social Science History Association Annual Conference’s Talk  (2005).
  -Geo-Recherche Conference’s Talk  (Winter 2005).
• Ftime - Fuzzy Time is an extension to PostgreSQL to store and manipulate imprecise (or fuzzy) moments and times in a relational database. Using fuzzy logic, this extension allows one to create, manipulate and query temporal information as expressed in natural language or in historical documents. FTime allows not only to model inaccurate dates and times such as "circa June 1845" or "during about two weeks" as an attribute in a database, but also to perform basic mathematical operations on this kind of data. It is for example possible to add or subtract "about five months" to "circa June 4, 1734" or to determine the duration between "circa July 12, 1765" and "circa September 1766". Ftime also allows finding moments in an imprecise manner. It is possible, for example, to select events that took place "around June 1765" and to sort them according to their degrees of relevance. See a PowerPoint presentation on Ftime  made at the CIEQ Annual Conference in May 2006.
• GeoWebSuite - GeoWebSuite is a file management and processing solution for centers distributing georeferenced data. Developed for the Center for Geographic Information and Statistics of the Laval University library, GeoWebSuite facilitates the management and processing of geospatial data from the moment they are downloaded until their webcast. It boosts the speed of data processing by facilitating conversion from one format to another, their diffusion through maps and their update. GeoWebSuite consists of three modules.
  
  GeoConverter is a batch converter of geospatial datasets. It supports conversion from various formats (Coverage, Mid/Mif, E00, Shapefile, MapInfo TAB, zipped or not) to shapefiles and TAB (zipped or not). GeoConverter allows unzip/convert/rezip, in a snap, hundreds of files and produce, in one step, their equivalents shapefile and Mapinfo versions, zipped or not. GeoConverter allows the insertion of personalized files or other metadata in each ZIP files.
  
  GeoIndex is a cartographic website developed with ArcIMS providing a comprehensive access to georeferenced file. The site allows one to select and download files through different layers of cartographic indexes. These indexes are superimposed on a base map that facilitates the location of files. GeoIndex see in action !
  
  GeoAccess is the essential little brother of GeoIndex. It allows updating and configuring GeoIndex index layers automatically when new files are to be posted. In one mouse click, GeoAccess update the indexes and the GeoIndex interface listing the new files available for each layer. No manual editing of the index, no change in the web page, no ArcIMS server restart: everything is fluid and automatic!
• GeoWebCampus  - GeoWebCampus is the campus of Laval University in 3D in Google Earth!

SCIENTIFIC PUBLICATIONS

Book chapters

• Urbano, F., Basille, M. and Racine. P. (2014) Chapter 6 - From Points to Habitat: Relating Environmental Information to GPS Positions.  In Spatial Database for GPS Wildlife Tracking Data. (Urbano, F. and Cagnacci, F., Eds.) Springer International Publishing, Switzerland.
• Basille, M., Urbano, F., Racine. P., Capecchi, V. and Cagnacci, F. (2014) Chapter 7 - Tracking Animals in a Dynamic Environment: Remote Sensing Image Time Series.  In Spatial Database for GPS Wildlife Tracking Data. (Urbano, F. and Cagnacci, F., Eds.) Springer International Publishing, Switzerland.
• Urbano, F., Basille, M. and Racine. P. (2014) Chapter 9 - Exploring Tracking Data: Representations, Methods and Tools in a Spatial Database.  In Spatial Database for GPS Wildlife Tracking Data. (Urbano, F. and Cagnacci, F., Eds.) Springer International Publishing, Switzerland.

Articles published in journals with revision committee

• Cumming, S., Drever, R., Houle, M., Cosco, J., Racine, P., Bayne, E., Schmiegelow, F.K.A. (2015) A gap analysis of tree species representation in the protected areas of the Canadian boreal forest: applying a new assemblage of digital Forest Resource Inventory data . Canadian Journal of Forest Research, 2015, 45(2): 163-173, 10.1139/cjfr-2014-0102.
• Raulier, F., Dhital, N., Racine, P., Tittler, R., Fall, A. (2014) Increasing resilience of timber supply: How a variable buffer stock of timber can efficiently reduce exposure to shortfalls caused by wildfires . Forest Policy and Economics, 46(9):47–55.
• Beguin, J., McIntire, E.J.B., Fortin, D., Cumming, S.G., Raulier, F., Racine, P., Dussault, C. (2013) Explaining Geographic Gradients in Winter Selection of Landscapes by Boreal Caribou with Implications under Global Changes in Eastern Canada . PLoS ONE, 8(10): e78510. doi:10.1371/journal.pone.0078510
• Boivin, F., Paquette, A., Racine, P. and Messier, C. (2011) A fast and reliable method for the delineation of tree crown outlines for the computation of crown openness values and other crown parameters . Canadian Journal of Forest Research, 41(9) 1827-1835, 10.1139/x11-107.

Other publications

• Houle, M., Cumming, S.G., Racine, P. Common Attribute Schema for Forest Resource Inventory (CASFRI): Technical Assessment for Integration within CFS to Support an “Integrated Systems Approach”. Report for ISA Secretariat, CFS, Natural Resources Canada, Ottawa. 27pp.

TALKS

• Introducing the PostGIS Add-ons: An easy way to add functionality to PostGIS , FOSS4G 2017 , Boston, août 2017.
• Beyond points - How to turn SMW into a complete Geographic Information System , Semantic Mediawiki Conference 2013 , New-York, USA, March 2013. Along with two short presentations: SMW Metaproperties - An alternative to Page Schema and XML to facilitate form management  and A provocative proposition - Rethinking SMW for Dummies .
• GeoCMSConnect - Un framework pour le développement facile d'extensions géo dans les systèmes de gestion de contenu , Vision Géomatique 2012 , Saguenay, Canada, octobre 2012.
• Quoi de neuf dans PostGIS 2.0? Le raster bien sûr! Mais beaucoup plus! , Vision Géomatique 2012 , Saguenay, Canada, octobre 2012.
• Building the Canadian Spatial Data Foundry - An online portal for large scale spatial analysis , Global Geospatial Conference 2012 , Québec, Canada, mai 2012.
• Stockage, manipulation et analyse de données matricielles avec PostGIS Raster , Géomatique 2011 , Montréal, Canada, octobre 2011. (Sur YouTube )
• Store, manipulate and analyze raster data within the PostgreSQL/PostGIS spatial database , FOSS4G, Denver , USA, septembre 2011.
• Stockage, manipulation et analyse de données matricielles avec PostGIS Raster , Session PostgreSQL, Paris, France, juin 2011.
• Automatisation de la conversion des inventaires forestiers Canadiens avec ArcGIS et Python , Conférence régionale des utilisateurs ESRI, Québec, avril 2011.
• L'analyse de données matricielles dans une base de données géospatiale: L'approche PostGIS Raster , Centre d’information topographique, Ressources naturelles Canada, Sherbrooke, janvier 2011.
• L'analyse de données matricielles dans une base de données géospatiale: L'approche PostGIS WKT Raster , Centre de recherche en géomatique, Département de géomatique, Université Laval, octobre 2010.
• Introducing PostGIS WKT Raster: Seamless Raster/Vector Operations in a Spatial Database , FOSS4G, Barcelone , Espagne, septembre 2010.
• The Canadian Spatial Data Foundry: Introduction to PostGIS WKT Raster and Raster Objects , Géomatique 2009, Montréal, octobre 2009.
• PostGIS WKT Raster Specifications Version 1.0 - Seamless operations between vector and raster layers , décembre 2008.
• Représentation d’itinéraires biographiques dans un système d’information géographique temporel , Séminaire de maîtrise, Université Laval, juin 2006.
• GéoIndex: un nouvel accès aux données géospatiales du Centre d’information géographique et statistique de la Bibliothèque de l’Université Laval , GeoTec, Ottawa, juin 2006.
• FTime: Un outil de manipulation des moments historiques imprécis dans les bases de données relationnelles à caractère historique , Colloque du CIEQ , Université Laval, mai 2006.
• ArcIMS, une solution avantageuse pour un accès privilégié aux données géospatiales du Centre d’information géographique et statistique de la Bibliothèque de l’Université Laval , Conférence régionale des utilisateurs ESRI, Québec, avril 2006.
• Mapping biographical trajectories of historical characters in space and time: a practical approach , (Social Science History Association (SSHA) Annual Conference, Portland, Oregon, 2005)
• Modélisation et cartographie de récits de voyages dans un système d’information géographique , Colloque Géorecherche, Université Laval, hiver 2005.

TUTORIALS

• Atelier analyse spatiale dans PostGIS en SQL .
• Comprendre les systèmes de coordonnées  en général mais surtout dans ArcGIS 9.x.
• Intersecting vector polygons with large raster coverage using PostGIS Raster