ANAPOLIS. Analysis of polygonal terrains on Mars based on Earth analogues
Funding organization:
Fundação para a Ciência e a Tecnologia
Period: 01-2010 / 06-2012
Host Institution
Instituto Superior Técnico (IST/UTL)
Participating Institutions
Centro de Estudos Geográficos, Universidade de Lisboa, Portugal
Faculdade de Ciências e Tecnologia da Universidade de Coimbra (FCT/UC)
The University Centre in Svalbard, Norway (UNIS)
Principal Investigator: Pedro Pina
Team members: Adriane Machado, Alexandre Trindade, Alice Ferreira, Carla Mora, Fátima Lira, Gonçalo Vieira, Hanne Christiansen, José Saraiva, Lourenço Bandeira, Marc Oliva, Marco Jorge, Mário Neves, Miguel Angel de Pablo, Pedro Pina, Teresa Barata.
Fundação para a Ciência e a Tecnologia
Period: 01-2010 / 06-2012
Host Institution
Instituto Superior Técnico (IST/UTL)
Participating Institutions
Centro de Estudos Geográficos, Universidade de Lisboa, Portugal
Faculdade de Ciências e Tecnologia da Universidade de Coimbra (FCT/UC)
The University Centre in Svalbard, Norway (UNIS)
Principal Investigator: Pedro Pina
Team members: Adriane Machado, Alexandre Trindade, Alice Ferreira, Carla Mora, Fátima Lira, Gonçalo Vieira, Hanne Christiansen, José Saraiva, Lourenço Bandeira, Marc Oliva, Marco Jorge, Mário Neves, Miguel Angel de Pablo, Pedro Pina, Teresa Barata.
Abstract
One of the most intriguing aspects of the surface of Mars is the extensive presence of polygonal networks. These are geomorphological features defined by a duality (an edge and an interior) that gives rise to a somewhat geometric arrangement on the ground. The contrast between the two components can be expressed topographically, in reflectance terms, or both. The appearance that these networks show on the surface of Mars can be extremely diverse, and their dimensions vary between the hundreds and less than five meters, though they are somewhat constant within a given network. Their distribution on the surface has been correlated with the presence of ice in the Martian soil, and the most accepted hypothesis for their presence takes into account their similarities to polygonal networks on the Earth, which occur in periglacial areas. However, there are still many questions that remain to be answered about these features.
One of the teams involved in this project has conducted research into the automated identification and characterization polygonal networks on Mars, achieving very good results. The areas studied contain thousands of polygons, thus ensuring that the results have statistical significance, and that the geometric and topological parameters extracted can serve the purpose of distinguishing between different types of network. The methodology used is based on the processing of images of the surface of the planet acquired by diverse probes and their cameras, at different spatial resolutions. The one aspect of this work that can be improved is related to the validation of the results, which on Mars is dependent on a visual, and thus subjective, interpretation of the orbital images.
The core of this project is the coupling of multi-scale remotely acquired images (subject to adequate processing), where the algorithm for the automatic extraction of the networks is applied, with detailed (differential global positioning techniques) geodetic field surveys, which provide precise ground truth. Once data on Martian networks is mostly geometry related, and high geometric diversity exists, the collection of geomorphological data for terrestrial polygonal networks with different kinds of arrangements is crucial to establish analogies between the features on the two planets.
Since polygonal networks on Mars exhibit a wide diversity in dimensions, locations, visual aspect and even topography, their classification into types may lead to some interesting clues about their age, precise mechanism of formation and evolution through time, including its relation with major climate changes that occur on the planet due to changes in orbital geometry. Furthermore, this can point to a clear separation between polygons of periglacial origin and others that may have tectonic origins but, on the strength of visual analysis only, cannot be precisely identified. Thus, analysis can be focused on those polygonal networks that are most strongly suspected to be related to the history of water on the planet, and whose terrestrial analogues will be studied in detail, from remote sensed images and on the field. The effort to classify the polygons into classes or types is based on the possibility of distinguishing between the conditions and detailed mechanisms involved in the origin, spatial arrangement and time evolution of the networks. It should be expectable that networks that exhibit similar geometric and topological characteristics share similarities in what concerns their formation. Thus, we plan to enlarge the set of parameters that have already been employed in an effort to discriminate between networks with diverse visual appearance; information on type of intersection, orientation of edges, 3D shape and (when possible) granulometry will be considered. Methods of multivariate data analysis will be used to try and find discriminating factors that can clearly separate different types of networks; this will then be related to other information on the environmental and geologic conditions of occurrence, so that clues to the precise mechanisms of formation and evolution of the networks can be found.
The full development of this project will lead to the completion of a robust methodology for the identification and characterization of polygonal networks on Mars, the Earth and any other planetary surfaces where similar features occur. This will surely permit some advances in our understanding of this type of features on Mars, their relation with current and past environmental conditions, but it will also constitute a very useful tool (filling a perceived gap in the methods employed) for the study of terrestrial periglacial polygons and what it can tell us about the changing conditions on our planet. Clues about genetic processes, precise mechanisms of formation and evolution, relation between morphological characteristics and environmental conditions can be found through the detailed analysis of the results of this project and the input of the diverse teams.
One of the most intriguing aspects of the surface of Mars is the extensive presence of polygonal networks. These are geomorphological features defined by a duality (an edge and an interior) that gives rise to a somewhat geometric arrangement on the ground. The contrast between the two components can be expressed topographically, in reflectance terms, or both. The appearance that these networks show on the surface of Mars can be extremely diverse, and their dimensions vary between the hundreds and less than five meters, though they are somewhat constant within a given network. Their distribution on the surface has been correlated with the presence of ice in the Martian soil, and the most accepted hypothesis for their presence takes into account their similarities to polygonal networks on the Earth, which occur in periglacial areas. However, there are still many questions that remain to be answered about these features.
One of the teams involved in this project has conducted research into the automated identification and characterization polygonal networks on Mars, achieving very good results. The areas studied contain thousands of polygons, thus ensuring that the results have statistical significance, and that the geometric and topological parameters extracted can serve the purpose of distinguishing between different types of network. The methodology used is based on the processing of images of the surface of the planet acquired by diverse probes and their cameras, at different spatial resolutions. The one aspect of this work that can be improved is related to the validation of the results, which on Mars is dependent on a visual, and thus subjective, interpretation of the orbital images.
The core of this project is the coupling of multi-scale remotely acquired images (subject to adequate processing), where the algorithm for the automatic extraction of the networks is applied, with detailed (differential global positioning techniques) geodetic field surveys, which provide precise ground truth. Once data on Martian networks is mostly geometry related, and high geometric diversity exists, the collection of geomorphological data for terrestrial polygonal networks with different kinds of arrangements is crucial to establish analogies between the features on the two planets.
Since polygonal networks on Mars exhibit a wide diversity in dimensions, locations, visual aspect and even topography, their classification into types may lead to some interesting clues about their age, precise mechanism of formation and evolution through time, including its relation with major climate changes that occur on the planet due to changes in orbital geometry. Furthermore, this can point to a clear separation between polygons of periglacial origin and others that may have tectonic origins but, on the strength of visual analysis only, cannot be precisely identified. Thus, analysis can be focused on those polygonal networks that are most strongly suspected to be related to the history of water on the planet, and whose terrestrial analogues will be studied in detail, from remote sensed images and on the field. The effort to classify the polygons into classes or types is based on the possibility of distinguishing between the conditions and detailed mechanisms involved in the origin, spatial arrangement and time evolution of the networks. It should be expectable that networks that exhibit similar geometric and topological characteristics share similarities in what concerns their formation. Thus, we plan to enlarge the set of parameters that have already been employed in an effort to discriminate between networks with diverse visual appearance; information on type of intersection, orientation of edges, 3D shape and (when possible) granulometry will be considered. Methods of multivariate data analysis will be used to try and find discriminating factors that can clearly separate different types of networks; this will then be related to other information on the environmental and geologic conditions of occurrence, so that clues to the precise mechanisms of formation and evolution of the networks can be found.
The full development of this project will lead to the completion of a robust methodology for the identification and characterization of polygonal networks on Mars, the Earth and any other planetary surfaces where similar features occur. This will surely permit some advances in our understanding of this type of features on Mars, their relation with current and past environmental conditions, but it will also constitute a very useful tool (filling a perceived gap in the methods employed) for the study of terrestrial periglacial polygons and what it can tell us about the changing conditions on our planet. Clues about genetic processes, precise mechanisms of formation and evolution, relation between morphological characteristics and environmental conditions can be found through the detailed analysis of the results of this project and the input of the diverse teams.
© Norwegian Polar Institute.
top left: Subset of aerial image S2009_13835_00142, top and bottom right: subset of aerial image S2009_13835_00144, bottom left: Subset of aerial image S2009_13835_00145
top left: Subset of aerial image S2009_13835_00142, top and bottom right: subset of aerial image S2009_13835_00144, bottom left: Subset of aerial image S2009_13835_00145
