Eduardo Ivo Alves

The present study focusses on a stony meteorite found in the 90’s in North West Africa, that has not yet been classified. It was studied by optical microscopy, X-ray fluorescence, X-ray diffraction and Mössbauer spectroscopy. The Fe-bearing phases found by Mössbauer spectroscopy are consistent with the elements found by XRF measurements as well as with phases determined by XRD patterns. The RT spectrum is complex mainly due to the large linewidths of the lorentzians or to the distribution of sites that can obscure the presence of sextets or doublets. From the results obtained the classification of the NWAXX meteorite can be made to an achondrite Winonaite.

Workshop Report on the International Meeting on Paleoclimate: change and adaptation (Coimbra, 2019) - Portugal

Foreword Climate is the planetary response of the atmospheric circulation to its changing composition, to the solar system configuration, to the Earth’s rotation and to the oceans’ and continents’ distributions. It displays, as a result, a restless moving pattern, expressed at a global scale by subsiding and uplifting convection cells. These changes have long been recognized and documented in geologic objects of all ages. In many rocks, different geologic features, fossil fragments and imprints, prehistoric remains and historic reports, there is a climate signal that can be analyzed and interpreted. All that information should be gathered in order to learn more about past climate changes. Lessons from the past support the view that deep change is the rule, not the exception, even where no reference is available, due to strongly contrasting extremes, chaotically defined by the whole ensemble of extra-planetary, external, and internal geodynamic controls. Science-based knowledge is crucial to face current challenges, which are focus for research within the UNESCO chair on Geoparks, Sustainable Regional Development and Healthy Lifestyles and the UNESCO chair on Humanities and Cultural Integrated Landscape Management, both partners in this international meeting, which has been organized by the Geosciences Centre and the Centre for Earth and Space Research of the University of Coimbra. This volume records the contribution of about seventy ongoing projects developed by research teams with a wide range of scientific backgrounds from different regions of the planet, who met at the University of Coimbra on the 18th-19th June 2019. The enriching discussions on paleoclimates in the Solar System, climate changes in geological time, climate memory in the geological record, climate changes and human adaptations throughout the Quaternary, and climatic events and human-environment interactions in the Holocene, indicate that the main objective underlying this initiative will have been fulfilled: to stimulate an observational attitude and to promote an open discussion on paleoclimatic signals in order to improve our look at the present and to ground future perspectives. The Organizing Committee International Meeting on Paleoclimate: Change and Adaptation University of Coimbra, 18th-19th June 2019

We present a simple method to enhance the view of Titan drainage networks, thus allowing extracting relevant hydrological parameters. The method uses RGB fusion of three Cassini synthetic aperture radar images acquired at different times, and is tested on one drainage network. Comparison with previous studies of the same network shows an increase in all the measured parameters. The present results help constrain previous estimates of erosion times, terrain, and tectonic models for the area and indicate that, whenever possible, geomorphological inference from drainage network geometry should be drawn on multiply sampled scenes.

Os derrames naturais de hidrocarbonetos (oil seeps) permitem frequentemente inferir a existência de uma rocha reservatório em profundidade e, como tal, a geração de hidrocarbonetos. A ocorrência de oil seeps é também a expressão da existência de caminhos de migração que permite a ascensão de hidrocarbonetos até à superfície. Em Portugal a geração de hidrocarbonetos é comprovada pela existência de três oil seeps onshore localizados em Leiria, Paredes de Vitória e Torres Vedras. O presente trabalho apresenta os resultados preliminares de um estudo em curso, que tem como objectivo identificar evidencias de oil seeps no offshore Português até agora desconhecidos. Este estudo tem como base a análise de imagens de satélite SAR (synthetic aperture radar) para identificação de manchas anómalas, tendo sido reconhecidos mais de uma centena de oil seeps em todo offshore Português. Este trabalho apresenta alguns resultados de um estudo de detecção remota, que consiste na análise de imagens de satélite SAR pelo satélite SENTINEL 1A. O registo destas imagens não é influenciado pelas condições atmosféricas, como a presença de nuvens ou pela luz solar. O maior desafio desta análise é o reconhecimento correcto de possíveis oil seeps, visto que em offshore ocorrem manchas anómalas com outras origens (look-alikes). Estas manchas estão associadas a actividade antrópica, biogénica ou a upwelling. Este estudo divide-se em duas etapas; (i) análise manual das imagens de satélite para reconhecimento inicial de manchas anómalas, com base no contraste de cor da mancha com o background; (ii) análise das imagens com base na utilização de um algoritmo de segmentação semi-automático (histerese) para uma determinação mais fiável dos oil seeps. As manchas anómalas identificadas foram comparadas com informação relativa ao padrão teórico dos oil seeps, para reduzir qualquer confusão com derrames de origem antrópica. Até hoje foram identificadas manualmente mais de uma centena de manchas anómalas, a partir da observação de 925 imagens SENTINEL, registadas de 2014 a 2016. Estas imagens estão agora a ser analisadas com recurso a um algoritmo, não só para delimitar de uma forma mais fiável o limite água-óleo, mas também para eliminar falsas manchas anómalas. Com base na observação manual foi possível constatar que em diversos locais ocorre esporadicamente derrame de hidrocarbonetos (Figura 1). O formato dos oil seeps pode ser diverso, devido à dispersão dos mesmos por acção das correntes oceânicas e do vento (Figura 1B-E). Estes possíveis oil seeps foram registados em locais muito próximos e em diferentes datas, de Maio a Setembro de 2015. O registo frequente de manchas no mesmo locais em momentos distintos implica que a ocorrência de derrames seja ocasional em offshore. Devido à localização geográfica destes oil seeps (Figura 2) é possível sugerir que estes derrames estejam associados à tectónica local, uma vez que ocorrem junto a grandes estruturas tectónicas. Estas estruturas poderão ser responsáveis pela migração de hidrocarbonetos, induzindo à sua ascensão desde o reservatório até ao fundo marinho, acumulando-se à superficie. A Figura 3 mostra o resultado da aplicação do algoritmo, verificando-se que esta análise permite uma eficaz delimitação do oil seep. Este processo corresponde à fase inicial da análise semi-automática das imagens de satélite, seguindo-se depois o estudo dos parâmetros estatísticos dos derrames, para assim proceder à classificação automática dos oil seeps. É conveniente referir que esta metodologia tem sido capaz de distinguir oil seeps de manchas biogénicas e de efeitos de corrente. A existência de oil seeps implica a geração de hidrocarbonetos em território português. Assim, torna-se necessário conhecer o(s) sistema(s) petrolífero(s) que estarão na origem dos oil seeps, e averiguar a sua possível relevância exploratória. Figura 1. Ilustração do conjunto de oil seeps reconhecidos com base na observação manual de imagens SENTINEL 1A, de Maio a Setembro de 2015 (A). Estes oil seeps ocorrem com a mesma localização geográfica mas em datas distintas, revelando que o derrame de hidrocarbonetos é frequente. (B-D) detalhe dos restantes oil seeps. Figura 2. Correlacção entre os oil seeps reconhecidos e das estruturas tetónicas de ruptura. Existe uma grande afinidade entre as estruturas assinaladas e os oil seeps reconhecidos, tornando evidente que o derrame de hidrocarbonetos associado a estas falhas é frequente. Figura 3. Detalhe de um oil seep (A). Detalhe do resultado da aplicação da histerese na mesma imagem (B).

Let there be Ai, any parameter of an actual earthquake, and Fi, the forecast that was made of the same parameter for the same earthquake, and N the number of (Ai, Fi) test pairs. These objectives were attained by integrating all prediction algorithms in a Microsoft™ Excel®workbook called NCEL. Worksheet “Main” is the only one where the user can input data: a series of up to 1000 records of four earthquake parameters, event time and magnitude, epicentral latitude and longitude. It should be stressed that input data must have some sort of pre-processing. In all the examples below, the series were truncated for the lower magnitude events (non-linear bottom part of the Gutenberg-Richter function). Also, as a crude rule-of-thumb method to filter off foreshocks and aftershocks only the strongest earthquake in each day was used. The user can also, if willing, hand-tune most of the parameters of the neural network: length of the training set, number of neurons in the middle layer, learning rate, learning momentum, limit target error and limit number of iterations. However, default values for all these parameters are supplied. Namely, in the earlier works, the length of the middle layer vector was chosen so that the total number of connections was of the same order of the number of training examples. Worksheet “Monte Carlo” performs Monte Carlo forecasts based on the input data on worksheet “Main” . Worksheet “Previous” performs naïve forecasts based on the input data on worksheet “Main” : it forecasts, for the next earthquake, exactly the same parameters that are recorded for the last one (which is a significantly successful method for weather forecasts for about half of the year). Worksheet “Averages”, another naïve approach (or maybe not so much, as will be seen below), forecasts, for the next event, the average for each parameter of all the previous events. Visual Basic module “NCEL” performs the neural network training and testing procedures. The training procedure is based on the algorithm provided by [19]. In its earlier versions, “NCEL” followed closely the previous method, with the exception of forecasting the epicentral co-ordinates. Several test runs made over well known datasets showed unexpected results: many times the forecasts that were made by the “Averages” naïve method produced the smallest errors, mainly in MAD. In order to quantify this result and to further automate the forecasting procedure, the current version of VB module “NCEL” was all included in a loop that stops on one of two conditions: a command that is entered by the user or [MAD(NN) < MAD(Averages)] for the time parameter. Worksheet “Error Analysis” keeps track of the number of times that the neural network performs better than the “Averages” algorithm. Why are the results not always the same? The “Previous” and “Averages” spreadsheets, once initially computed are, of course, invariant. Differences arise on each new calculation of spreadsheets “Monte Carlo” (because of the intrinsically quasi-random nature of the method) and “NN”. In the NN case, the differences occur because of the randomly initiated connections’ weights. However, unlike in “Monte Carlo” , the final results of different network runs never differ by more than 5% (on 300 tested cases).

In 1994, a new earthquake forecasting method was developed, that integrates in a neural network several numerical forecasting tools that had been originally developed for financial analysis. This method was fine-tuned and tested with the seismicity of the Azores (North Atlantic Ocean), predicting the July 1998 (M=6.2) and the January 2004 (M=4.3) earthquakes, and Continental Portugal, predicting the December 2005 (M=4.0) event. In no case were there false alarms. The same methodology was applied to seismicity along the Sunda Megathrust. The training set consisted of all recorded M5.5 events occurred since January 1900 with epicenters between latitudes 10 °S and 15 °N, and longitudes 90 °E and 105 °E, recorded at the United States Geological Survey earthquake database. After training, testing, and parameter tuning, the neural network forecast an earthquake for April 2015 (±5 months), with magnitude 5.5 (±1.0), with epicenter at coordinates 1 °S (±9°); 98 °E (±6°) with 95% confidence. That earthquake eventually occurred, on March 3, 2015, with magnitude 6.1, with epicenter at coordinates 0.779°S; 98.716°E.

Wrinkle ridges constitute one of the most abundant tectonic features on terrestrial planetary surfaces. On Venus, evidence suggests a connection between wrinkle ridges and the climatic evolution of the planet. However, like other planets and moons that experience more active surface geological processes, such as Earth, Mars, Europa, Io and Titan, visible impact craters on the Venusian surface are less common because they are eroded, buried or transformed by tectonics or other geological processes over time. It is of great importance to identify and understand some characteristics of those surface morphologies, such as orientation, length, spacing, original dimension and topography. Nevertheless, these parameters can only be computed on remotely sensed images after their segmentation. Until now, the manual identification of these features has been focused on those of major geological significance, leaving many more to be identified, mapped and studied. The main aim of this paper is to provide a method for automatic detection of wrinkle ridges from Magellan Synthetic Aperture Radar (SAR) imagery at different scales. The proposed algorithm, based on a combination of fractal dimension and morphological operators, identifies regions of interest to this study, namely those of anisotropic behaviour, but also impact craters and their ejecta blankets. The high performances achieved in a variety of situations demonstrate that its robustness can be applied to an automated procedure.

Earthquake prediction is usually dealt with in one of two ways: the prediction of the effects of a postulated earthquake on a specific site or the prediction of the occurrence of events. The first approach is today well developed and should always be instrumental for the design and implementation of construction codes and for general planning. Event forecasting still is regarded as a scientific curiosity without much practical application, namely where it still must prove to be useful: for the protection of human lives and property. This is the core question: can this type of forecast be useful? In the present work an application of earthquake forecasting is proposed that is similar in form to weather probability forecasting and should be made available both to the authorities and to the public. In the present context, the question that must be answered is: can this type of forecast be useful (in the above sense of helping to protect lives and property)? The prediction of another type of chaotic phenomenon has entered our lives for decades-weather forecasts-and perhaps we can draw lessons from its application. The sociological argument that earthquake forecasting can cause widespread panic has been surpassed by meteorology and climatology. A similar approach to earthquake forecasting is proposed here: the existing methods should be integrated in the periodical publication of seismic probability maps. This will have two beneficial effects: to get the general public acquainted with earthquake forecasting, its successes and failures, and, at the same time, helping to divulge a much needed culture of safety and preparedness; to the authorities, since civil protection resources are always scarce, the knowledge of an increased probability of an earthquake occurrence could help cluster some of those resources on the most threatened areas during the highest probability windows.

Room temperature 57Fe Mössbauer spectroscopy has been used to obtain information on the redox and coordination of iron in tektites. A MIMOS II spectrometer in backscattering geometry has been used in the study, so that no sample preparation at all was required. X-ray fluorescence has been used to determine the composition of the tektites. Mössbauer spectra have been deconvoluted using three extended Voigt-based profiles to allow quantitative analysis of iron atoms valence and coordination. In all tektites, the Fe2+sites have been distinguished in Fe with octahedral and tetrahedral coordination. The Fe2+octahedral sites show a region of isomer shift (IS) and quadrupole splitting (QS), IS = 1.02-1.14 mm/s and QS = 1.82-2.12 mm/s, relative to α-Fe. The Fe2+tetrahedral sites show a region of hyperfine parameters of IS = 0.59-0.89 mm/s and QS = 1.14-1.60 mm/s. The Fe3+sites show IS = 0.11-0.33 mm/s and QS = 0.02-0.04 mm/s. The Fe3+/Fe2+ ratio was found to be 0.025-0.149.

We present an assessment of methods used for the structural mapping of planetary surfaces. We compared two photo-interpretations and a dataset obtained semi-automatically.

Tektites are natural glasses originated from fused material spilled during a meteorite impact.

Room temperature 57Fe Mössbauer effect spectroscopy has been used to investigate the local Fe environment in a set of indochinite tektites from Thailand. A MIMOS II spectrometer in backscattering geometry has been used in the study, so that no sample preparation at all was required. The spectra have been analysed in terms of discrete spectral components using Voigt functions. The results are similar to those obtained with analysis done with quadrupole splitting distributions. In all cases the Mössbauer spectra show a broadened asymmetric quadrupole split doublet. The Fe 2+ sites have been distinguished in Fe with octahedral and tetrahedral coordination. The Fe3+ /Fe2+ and Fe2+ tetrahedral/octahedral ratios have been determined from the relative areas of the various spectral components. Their values have been discussed in the light of actual literature.

The presence of patterned grounds on Mars has been reported in several papers, especially the study of polygons distribution, size and formation processes. In the last years, the presence of basketball terrains has been noticed on Mars. Studies were made to recognize these terrains on Mars through the analysis of Mars Orbiter Camera (MOC) images. We have been developing an algorithm that recognizes automatically and extracts the hummocky patterns on Mars related to landforms generated by freeze–thaw cycles such as mud boils features. The algorithm is based on remote sensing data that establishes a comparison between the hummocks and mud boils morphology and size from Adventdalen at Longyearbyen (Svalbard—Norway) and hummocky patterns on Mars using High Resolution Imaging Science Experiment (HiRISE) imagery.

Palimpsest, ghost, or degraded craters have so far been identified on Mercury, the Earth, the Moon, Mars, Ganymede, Callisto, and possibly even Titan. We have identified at least 10 palimpsests on HiRISE image ESP_016526_2415 of Mars and developed an algorithm to automatically characterize these structures on high resolution images. Since the genesis of palimpsests may be related to freeze–thaw cycles these features might have provided habitable environments.

https://dl.dropboxusercontent.com/u/9273555/papers/custodioetal_srl2012.pdf The seismic station COI, located in Coimbra, Portugal, is the oldest seismic station in Portugal and among the oldest in the world. Experimental seismic recording began in 1891 with an Angot seismograph (V. G. S. S. Santos, unpublished data, 1995; Custódio et al., 2010). Continuous recording of ground motion began in 1903 with a Milne seismograph, and is still on-going at the same site, now with a broadband Streckeisen STS-2 seismometer. Coimbra’s observatory owns an over 100-year long, well-preserved collection of seismograms, seis- mographs and bulletins. Station COI is taken into account in early seismological works, for example, in the 1901 compila- tion of stations equipped with seismographs recommended by the British Association for the Advancement of Science (BAAS, 1901), in the earthquake catalog of the International Association of Seismology for the year of 1904 (Rosenthal, 1907), in the San Francisco earthquake report (Reid, 1910) or in the classical book “Seismicity of the Earth” (Gutenberg and Richter, 1954). However, throughout time station COI lost its early importance and was progressively forgotten. Station COI is missed even by recent compilations of his- torical seismic data (e.g., Ferrari, 2000; Schweitzer and Lee, 2003). Recently we were able to start undusting the seismological heritage of Coimbra’s station. In this paper, we present a brief history of the station. We also describe its seismographs, seis- mograms and bulletins, including their present state of preser- vation and availability. Station COI is particularly relevant for studies of earthquakes in the pre-digital era given that: (1) it is one of the westernmost stations in Europe (Fig. 1); (2) it has operated for more than one century at the same site; and (3) the instruments are located inside an underground shelter with double walls, on well-consolidated ground, in a site favor- able for seismic observations.

The best Titan surface resolution images available were acquired on January 14, 2005, in the descent trajectory of the Huygens probe. This work shows a synthetic resolution enhancement of a small drainage network, based on two Huygens images integrated in a super-resolution algorithm. (Article in Portuguese)

Two Portuguese research groups together with one Italian group will develop a methodology to automatically delineate and extract dimensional and morphological features of dune fields in images of the martian surface.

We report the finding of palimpsest craters on Mars HiRISE image ESP_016526_2415. These craters show clear relationships with periglacial features: polygonal terrains whose characteristics are controlled by the craters.

The present work aims to identify hummocks and putative mud boils on martian hummocky terrains based on Earth analogues from Adventdalen Valley, Longyearbean, Spitsbergen Island, Svalbard.

An automatic method for strain assessment was applied to a rift located in Thaumasia Planum. A transtensive regime for the rift formation was confirmed. The presented strain analysis gives a better perspective of the kinematics of the rift.

The present work is focused on a methodology for the detection of craters and ejecta deposits on Magellan SAR images of the surface of Venus, based on fractal dimension and morphological transforms.

RESUMO: Apresenta-se aqui uma base de dados, construída numa folha de cálculo comum, que permite localizar imagens de Vénus, obtidas pela missão Magalhães, a partir das suas características técnicas, como cobertura, resolução ou direcção de vista, das suas coordenadas, ou de nomes topográficos. A base de dados permite, ainda, localizar áreas no planeta que foram amostradas mais que uma vez. Esta capacidade pode ser útil em estudos de geodinâmica externa de curto prazo ou na super-resolução de imagens. PALAVRAS-CHAVE: Vénus, Magalhães, imagens, base de dados. ABSTRACT: We present a database, built on common spreadsheet software, which allows locating Venus Magellan images based on technical criteria, such as coverage, resolution or sight direction, on their coordinates or topographic names. The database also allows locating areas which were sampled more than once. This ability can be useful on short-time-range geodynamic studies or on the super-resolution of images.

RESUMO: Apresentam-se neste trabalho duas novas técnicas para cartografar automaticamente as deformações crustais associadas a dois tipos de estruturas tectónicas presentes na superfície de Marte: falhas normais e cristas de enrugamento (wrinkle ridges). Estas técnicas permitem a caracterização espacial dos padrões de deformação quer para estruturas compressivas quer distensivas. PALAVRAS-CHAVE: Análise automática de deformação, tectónica, Marte. ABSTRACT: Two new techniques for automatic mapping of crustal strains on Mars surface are presented. The strains related with normal faults and wrinkle ridges are analysed allowing the spatial characterization of the strain patterns associated with compressive and distensive structures.

The parameters of wrinkle ridges (length, size, orientation) are determinate in a simple way if the wrinkle ridges are easily detected. This work presents the preliminary results of automatic detection of wrinkle ridges from SAR imagery at different scales.

1] This study focuses on the volcanism in Syria Planum, located at the center of the Tharsis bulge at an altitude of 6 to 8 km above Mars datum. Syria Planum was previously recognized as a center for the tectonic activity of Tharsis, but not as a major locus for volcanic activity, despite its centrality over the bulge. Using high-resolution images from the high resolution stereo camera on Mars Express combined with Mars Observer Laser Altimeter data, we have characterized a volcanic system that reveals a number of very interesting aspects of Mars volcanism. We identified a swarm of tens of coalesced shallow volcanic edifices, typically 10–30 km diameter, 0.1–0.2 km high, and with slopes around 0.5°. These characteristics are similar to those of small shield volcanoes found in Iceland. In addition, an intermediate-sized volcano, which is the source of lava flows that extend over >200 km, is observed west of this shield swarm. Our study characterizes a previously unrecognized volcanic assemblage on Mars which appears to be much more developed than was documented before, in terms of morphology, inferred origin, and periodicity of eruption. The estimated lava flux of the Syria Planum volcanoes is of the same order as the lava flux of Tharsis Montes. These characteristics suggest that Syria Planum experienced a very specific style of volcanism, which we dated to the Hesperian period.

A method for automatic lineament extraction using topographic data is applied on the Thaumasia plateau. A comparison is made between the results that are obtained from the automatic mapping approach and from a traditional tectonic lineament mapping.

Here we present the first results achieved with the project SIGN (Signatures of environmental change in the observations of the Geophysical Institutes). The project’s main goal is to convert into a digital database the historical meteorological data, recorded since 1856 until 1940 in several annales published by the 3 Portuguese Geophysical Institutes (of Lisbon, Porto and Coimbra) and the Portuguese Meteorology Institute. The different sets of historical data contain monthly, daily and sometimes hourly records of pressure, temperature, precipitation, humidity, wind speed and direction, cloud cover, evaporation and ozone. The published data cover several stations in mainland Portugal, the Azores and Madeira islands and in former Portuguese African and Asian colonies. The main objective is to use the data to study the changes that have taken place in the historical records during the last 150 years, when the recovered data is merged with the post-1941 data stored in the Meteorology Institute digital database. The other aim is to make the data available to the meteorology community at large. Direct observations of pressure data for Lisbon in the 1856-1940 period were prioritised and have been manually digitised, being later subjected to quality control tests. Digital historical records of Lisbon temperature, relative humidity and precipitation data have been obtained through corrected OCR techniques applied to published hourly or bi-hourly tables. Preliminary digital results are also available for several stations in mainland Portugal, Azores and Madeira. Data for the Escola Médico-Cirúrgica station in Porto during the 1861-1898 period are already in digital format. All datasets are subjected to an initial quality control test, to detect wrong values, with more comprehensive tests to be applied at later stages. At the same time, detailed metadata files are being compiled for each station. This work will show the preliminary analysis results for the digital historical database obtained so far.

Here we present the first results achieved with the project SIGN (Signatures of environmental change in the observations of the Geophysical Institutes). The project’s main goal is to convert into a digital database the historical meteorological data, recorded since 1856 until 1940 in several annales published by the 3 Portuguese Geophysical Institutes (of Lisbon, Porto and Coimbra) and the Portuguese Meteorology Institute. The different sets of historical data contain monthly, daily and sometimes hourly records of pressure, temperature, precipitation, humidity, wind speed and direction, cloud cover, evaporation and ozone. The published data cover several stations in mainland Portugal, the Azores and Madeira islands and in former Portuguese African and Asian colonies. The main objective is to use the data to study the changes that have taken place in the historical records during the last 150 years, when the recovered data is merged with the post-1941 data stored in the Meteorology Institute digital database. The other aim is to make the data available to the meteorology community at large. Direct observations of pressure data for Lisbon in the 1856-1940 period were prioritised and have been manually digitised, being later subjected to quality control tests. Digital historical records of Lisbon temperature, relative humidity and precipitation data have been obtained through corrected OCR techniques applied to published hourly or bi-hourly tables. Preliminary digital results are also available for several stations in mainland Portugal, Azores and Madeira. Data for the Escola Médico-Cirúrgica station in Porto during the 1861-1898 period are already in digital format. All datasets are subjected to an initial quality control test, to detect wrong values, with more comprehensive tests to be applied at later stages. At the same time, detailed metadata files are being compiled for each station. This work will show the preliminary analysis results for the digital historical database obtained so far.

4) Instituto Geofísico da Universidade de Coimbra, Portugal, (5) Centro de Geofísica da Universidade de Lisboa, IDL, Lisbon, Portugal (mavalente@fc.ul.pt / Fax: +351 213908187 / Phone: +351 213921863) This work presents the joint efforts of the 3 Portuguese Geophysical Institutes (of Lisbon, Oporto and Coimbra) and the Portuguese Meteorology Institute to convert to a digital database the historical meteorology data, recorded since 1856 until 1940 in several annales published by the institutes. The different sets of historical data con-tain monthly, daily and sometimes hourly records of pressure, temperature, precipi-tation, humidity, wind speed and direction, cloud cover, evaporation and ozone. The published data cover several stations in mainland Portugal, the Azores and Madeira islands and in former Portuguese African and Asian colonies. One of the aims is to use the data to study the changes that have taken place in the historical records during the last 150 years, when the recovered data are joined with the post-1941 data stored in the Meteorology Institute digital database. The other aim is to make the data avail-able to the meteorology community at large. Direct observations of pressure data for Lisbon and for the 1856-1940 period were prioritised and have been manually digi-tised, being later subjected to quality control tests. Digital historical records of Lisbon temperature, relative humidity and precipitation data have been obtained through cor-rected OCR techniques applied to published hourly or bi-hourly tables. Preliminary digital results are also available for several stations in mainland Portugal, Azores and Madeira. All datasets are subjected to an initial quality control test, to detect wrong values, with more comprehensive tests to be applied at later stages. At the same time, detailed metadata files are being compiled for each station. This work will show the

Mars Image Mining System (MIMS) is a relational database of internet links to most of the images which were acquired by missions Mariner 9, Viking 1 and 2, Mars Global Surveyor, Mars Odyssey and Mars Express (a total of 329,240 images on December 15, 2005). Unlike existing online databases, it allows to search for images of several missions simultaneously, based on criteria such as image resolution, location and nearest feature name. Also, MIMS allows searching for images acquired on the same location by different missions, thus enabling the user to evaluate the evolution in time (from 1971 to 2005) of external geodynamic phenomena and features such as dust coverage, dune fields, the formation of gullies or cratering.

In 1994, a new earthquake forecasting method was developed, that integrated in a neural network several forecasting tools that had been originally developed for financial analysis. This method was tested with the seismicity of the Azores, predicting the July, 1998, and the January, 2004, earthquakes, albeit within very wide time and location windows. Work is beginning to integrate physical precursors in the neural network, in order to narrow the forecasting windows

Wavelet analysis and mathematical morphology are applied to MOLA altimetry data to automatically extract tectonic features from Mars' surface.

The surface of Mars is crossed by many chains of grabens that are mainly disposed radially from the centre of Tharsis volcano-tectonic province. The characterisation of these grabens has been done by visual analysis and interpretation of remote sensing imagery. More recently the use of MOLA (Mars Orbiter Laser Altimeter) DEM (digital elevation models) has allowed a better geometric characterisation of the fault scarps. We propose the use of an automated and quantitative methodology, based on wavelet analysis and mathematical morphology transforms, which permits the identification and segmentation of fault scarps. We apply several digital image processing techniques on MEGDR (Mission Experimental Gridded Data Records) in order to extract the fracturation pattern.

This paper presents a methodology to automatically recognise impact craters on the surface of Mars. It consists of three main phases: in the first one the images are segmented through a PCA of statistical texture measures, followed by the enhancement of the selected contours; in a second phase craters are recognised through a template matching approach; in a third phase the rims of the plotted craters are locally fitted through the watershed transform.

We intend to show that there is a negative correlation between areas of magnetic anomalies and areas of energetic particles bombardment on the surface of Mars, by comparing MGS MAG-ER and Mars Odyssey MARIE maps. Terra Sirenum is the most shielded area.

There is consensus that seepage of fluids occurred through Martian layers, with formation of channels and gullies that are much similar to those on Earth. However, it is not yet firmly established whether the seeping fluid was liquid water or not. Also, the time frame of these events is undefined. In the present work we address this latter issue, by comparing images that were acquired in two different missions: Mars Global Surveyor (MGS) and Mars Odyssey (MO). A comparison of image V01138003, obtained by the THEMIS camera onboard the MO spacecraft on March 18, 2002 (1) and image MOC2-240, obtained by the Mars Orbiter Camera onboard the MGS spacecraft on July 12, 1999 (2), shows a new gully on the northern scarp of Nirgal Vallis, near 29.7°S, 321.4°E, among previously existing gullies, with discernible formation of an incipient fan-shaped apron. This shows that fluid outflow is happening in the present. The fact that the THEMIS image is of medium-resolution (18 m/pixel) does not allow seeing many details on the new gully. This spot is consequently a good candidate for image acquisition by the Mars Express HRSC camera that will have a maximum resolution of 2.3 m/pixel, with the further advantage of stereoscopic vision. 1. Arizona State University 2. Malin Space Science Systems

Cratering is the dominant external geodynamic process on terrestrial planets. Crater counting is an essential step in planetary geologic studies, albeit a heavily time-consuming and manpower-consuming one. The present work introduces an automatic crater recognition method that compares favourably with other approaches, such as the circular Hough transform: it is less sensitive to noise, it is less memory-demanding and it is well suited to fast implementations that can be the base for autonomous landing guidance systems on very remote targets. The method is very simple, being based on classic image enhancement and recognition procedures: A. Image pre-processing. 1.Original greyscale image. 2.Stretch histogram between 0 and 256. 3.3x3 mean filter. 4.Detect edges with Sobel convolution. 5.Threshold at average + 2 standard deviations, producing a 2-bit image. B. Crater recognition (on 2-bit image). 1.Input Rmin, Rmax and Arc - minimum and maximum search radiuses, and circle arc to be considered a crater (between 1 and 2 PI). 2.R=Rmax. 3.x=x0; y=y0 4.Count all 1-pixels that are one radius away from (x, y). 5.If number of pixels is greater than Arc, then there is a crater of centre (x, y) and radius R. 6.Erase pixels on recognised crater (make every recognised 1-pixel a 0-pixel). 7.Increase centre coordinates until all pixels are covered. 8.Decrease R by one pixel until Rmin is reached. On all subsequent tests, Rmin=2 (pixel), Rmax=Image_side/4, [1.00PI

We show how magnetic gradiometry can increase our knowledge of the planet's life sustainability. It can also help us select a landing site for the first human explorers and increase our knowledge of the geology and geological evolution of Mars.

After the pioneering works of Beno Gutenberg and Charles F. Richter, not many at- tempts have been made to quantify world seismicity as a whole. The problem today is twofold. First, there is the issue of the amount of data that is available. Then there is the issue of the usefulness of this analysis: is there such a thing as a "global seismicity"? The present work managed the first issue by analysing only earthquakes of magnitude M>4, in the 25 years period between 1976 and 2000, a total of 39035 events from the NEIC catalogue. This truncated time-series adjusts well to a Gutenberg-Richter frequency-magnitude relation with a=9.88 and b=1.07, with R=0.99. The question of the existence of a global seismicity - that is, if there are long-ranged tectonic interactions on a global scale - must, however, be answered by other means. A first approach was to compute the Omori power law for aftershock decay rate parameter p for the 13 strongest earthquakes in the series (...)

Three Portuguese groups responded to the European Space Agency (ESA) second an- nouncement of opportunities for the Mars Express Mission. Projects were submitted in the fields Atmosphere and Surface/Atmosphere Interaction, Exobiology and Geologi- cal Evolution. Contacts among the groups revealed large areas of potential interaction and a necessity of active data, methods and analysis interchange. The projects will be executed in the period 2002-2006, and aim to analyse the information that will be conveyed by instruments aboard the Mars Express Orbiter and Beagle 2 Lander. Our combined projects propose to yield seven products: the Mars Orbital Viewer U MOVie U ; an apparent thermal inertia (ATI) map; a high-resolution water stability map; a lithostructural map; a hydrological potential map; a local assessment of poten- tial biomarkers; a geological map of Mars. MOVie will be a virtual-reality addressable map of Mars, based on stereoscopic images from the HSRC camera. The ATI map will draw on data from both the OMEGA (VNIR band) and PFS (NIR/TIR band) instru- ments. This map will enrich the existing atmospheric circulation models and help in the discrimination of lithotypes and hydric potential. One of our main objectives is to address the question whether water release does occur in the present epoch, which will be done by analysing the water vapour content in the low atmosphere. Data ar- riving from Beagle 2 will also be the basis for the assessment of methane contents on low martian atmosphere and soil-adsorbed gases. Atmosphere, soil and rock compo- sition data from the lander will allow for the correction of remotely sensed spectra. Classification will proceed in two phases: first an expert system will evaluate point lithologies from mineral and rock spectra; then those point lithologies will be mapped onto the planetary surface by a self-organising neural network. Mathematical mor- phology operators will also be used to perform the classification of Mars surface by incorporating textural features, thus biasing the neural network into defining litho- logical patterns that correspond to regions with a higher hydric potential. The hydric potential map will draw on data from the lithostructural map, the water stability map and the MARSIS - microwave - instrument. Integration of all data will be achieved on the final geological map of Mars. The procedures developed in these projects will be reusable, with minor adjustements due to different instrumental setups, on other planetary missions, even for remote sensing of the Earth.

One of the most distinctive characteristics of Mars is the strong dichotomy between the northern and southern hemispheres. This dichotomy is shown on features such as climate, topography, cratering, lithology, volcanism and crustal magnetization and is still not fully understood. The asymmetry of cratering, namely, has been interpreted as a result of endogenous factors, such as crustal thickness, lava flows and lithological composition of impact sites, or exogenous factors, such as the obliteration of craters by dust storms. The present work intends to help discriminate which of the two sorces, endogenous or exogenous, is the strongest control for crater distribution. To that pur- pose, we have measured the fractal dimensions of cratering and crustal magnetization in the same broad domains, namely the northern and southern hemispheres and the whole planet. The dimensions that were determined are all unifractal and comparable: there is less than 5% difference in dimension between cratering and magnetization in all the studied cases. Those dimensions were all low (between 1.2 and 1.5), indicating some linearity in the distributions. These data point to endogenous phenomena as the main controls for the present crater distribution asymmetry.

New data and a new interpretation of the Fundao granitic pluton are given. The new model is based upon field, petrographic, mineralogical and geochemical studies. Two zonations are suggested: 1) concentric, centrifugal, which originated by magmatic differentiation, and 2) excentric, centripetal, indicating hydrothermal alteration. Two major alignments (E-W sheared dykes and N 35o E faults) control the zonality. The history proposed is that the pluton was intruded approx 315 m.y. ago; dyke injection soon followed, then tectonic shearing and a further, post-tectonic intrusion of granite at approx 295 m.y. Late Hercynian faulting with hydrothermal activity took place at approx 285 m.y., related to post-tectonic endomagmatic hydrothermalism.-R.E.S.