SIMULATION OF THE 1755 TSUNAMI FLOODING AREA IN THE ALGARVE (SOUTHERN PORTUGAL): THE CASE-STUDY OF PORTIMÃO

The November 1st 1755 earthquake remains the most powerful and destructive to hit Europe so far. Although frequently associated with the city of Lisbon, this earthquake caused similar or greater damage and casualties in the southwest of the Algarve, where the seismic intensity was estimated at IX-X Mercalli Intensity Scale. Some minutes later, a tsunami increased the dimension of the disaster. Using simple techniques of simulation, we estimated the potential impacts of the occurrence of a similar event to the 1755 tsunami nowadays in Portimão.


Introduction
The 1755 earthquake, which reached a magnitude of 8.5, remains the most powerful and destructive to hit Europe so far. Within minutes, many lives were lost, populations displaced, livelihoods, homes and infrastructures were destroyed.
Shortly after the earthquake, a tsunami increased the death toll and the amount of damage. The tsunami hit both coasts of the North Atlantic. It reached the London harbour and it is reported from Norway, as well as from the African coast. However, the more destructive damage occurred in the Portuguese coast, south from Lisbon, in the Moroccan coast and in the Gulf of Cadiz. The downtown of Lisbon was hit by waves 6 meters high and at Cape São Vicente (Southwest of Portugal) the run-up height, evaluated from historical data, was greater than 15 meters . The data from Spain and Morocco reported waves greater than 10 meters high. In the Spanish coastal towns of Huelva and Cadiz the tsunami was also violent and the greatest part of the estimated 2,000 victims was due to the tsunami, which was known as the "maremoto de Cádiz" (MARTINEZ SOLARES and LÓPEZ ARROYO, 2004).
The 1755 earthquake marked a significant milestone in the way Portuguese society perceived not only this type of geological phenomenon but also the subsequent disaster response. Throughout Europe, the disaster was vigorously explained by Enlightenment philosophers as a natural phenomenon that could be scientifically understood rather than an Act of God that served only to punish the sinful. In contrast, a deeply religious Portuguese society was willing to consider the earthquake as God's punishment. Facing stiff social opposition to mitigate this perceived punishment Prime Minister Sebastião José de Carvalho e Melo, later made Marquis of Pombal, launched a disaster response operation that was extremely complex for the time. The operation was set up despite the lack of socio-political experience in dealing with such an unprecedented disaster and the non existence of any previous emergency planning. Although his priorities were to bury the dead in the face of threatening epidemics and to take care of the living by means of food and shelter, the Marquis of Pombal preserved public order, allocated a workforce for reconstruction, and relaunched the Portuguese economy both at the microeconomic and macroeconomic scales. The reconstruction effort followed an assessment of several possible reconstruction models and the search for structural measures of hazard mitigation.
Although frequently associated with the city of Lisbon the 1755 earthquake caused similar or greater damage and casualties, in the southwest of the Algarve, where the seismic intensity was estimated IX-X Mercalli Intensity Scale (SOUSA, 1919;BAPTISTA et al., 1998;CHESTER, 2001).
It is our purpose to delimitate the area flooded by the 1755 tsunami in the municipality of Portimão, which was one of the most devastated in the Algarve, and to assess the impacts of the occurrence of a similar event nowadays. The study area corresponds to a coastal strip well delimited, extending from river Alvor to river Arade. It is characterised by an average altitude of 25 meters high, the highest point being at 71 meters, displaying gentle slopes, with the exception of some cliffs in the area of Praia da Rocha (Figure 1).
We want to show that the determination of the flooded area is an important instrument not only in terms of disaster preparedness but also for the integration of mitigation measures in the strategy of development and land planning of the coastal area.

Data and methodology
Taking into consideration that the earthquake occurred in 1755, the devastation is well documented in engravings, reports in newspapers of the time, and letters between several distinguished people. Prime Minister Melo ordered a survey to be answered by the priests of all parishes of the country, known as the Marquis of Pombal Survey, which contributed immensely to a scientific analysis of the disaster response. In fact, this survey is extremely scientific considering it was written in the 18th century. From it we can identify the moment the earthquake started and how long it lasted as well as the extension of the area affected; characteristics of the event and physical impacts including aftershocks, direction of the waves, disaster response and recovery efforts. Two of the questions focused specifically on tsunami 1 .This was not only an important intervention at the political level but it also was a significant initiative from a scientific perspective. The observation network that was thus established to document the effects of the earthquake was therefore extremely extensive and specific (PEDROSA and RICHARDS, 2005).
Another fundamental source is the Dicionário Geográfico [Geographical Dictionary] which contains vital information regarding several aspects of the geography of the Kingdom of Portugal in the 18th century. Although the process of collecting information by sending the survey to the priests, of all the parishes, began in 1721, several problems made it necessary to send a new survey in 1758 (COSTA et al. 2005). The questionnaire included a question 2 regarding the devastation level caused by the earthquake and the state of current reconstruction. Despite the fact that it did not include a question directly related with the tsunami, the priests mentioned, in their answers, the damage caused by the phenomenon. This survey is much more limited for the study of the 1755 earthquake than the earlier Marquis of Pombal Survey, but the answers to the latest regarding the South of Portugal remain unknown. For this reason the main support of our research is the Dicionário Geográfico of 1758.
The information collected in historical documents about the height of the waves and the extension of penetration of the tsunami on land provided a base to the delimitation of the flooded area. However, we are not entirely sure of the exact coastal morphology of the time, so it is particularly important to determine the location of churches and other buildings damaged by the force of the waves, inside of which the maximum height of the water is known, in order to validate our methodology. A good example is the Igreja da Misericórdia, located ten meters above sea level. The water, inside this building, is said to have reached 2.64 meters high. So the height of the wave was determined based in three types of information: historical description of the wave height, extension of penetration of the tsunami and reports of destroyed buildings.
Also important was the Digital Terrain Mode (DTM) built for the municipality of Portimão. The determination of the flooded area depends on the accuracy of this model. The DTM was built based in topographic information at 1/5000 scale, resulting in a Triangulated Irregular Network.
After determining the probable height reached by the wave in certain areas of the municipality, we used the analysis capacity of Geographical Information Systems technology to determine the area likely to have been flooded in the 1755 tsunami. The next step was cross-referencing this information with the current data available regarding population, built-up area, and equipments using overlay operations in order to evaluate the impacts of a tsunami similar to the one of 1755, nowadays.

Seismic hazard in the Algarve
In Portugal the Algarve is one of the regions most susceptible to the occurrence of earthquakes. The first known earthquake to have been felt in the Algarve dates back from 63 B.C. (OLIVEIRA, 1986), but over the centuries other events more or less intense have occurred in that area (Table 1).
In the 18th century three earthquakes hit the Algarve causing great destruction. The 1719 earthquake had its epicentre off the coast of Portimão and reached an estimated magnitude of 7. The 1722 earthquake, with a probable epicentre off the city of Tavira caused great damage in the villages of Portimão, Albufeira, Loulé and in the cities of Faro and Tavira causing many deaths and the ruin of churches, convents, wall towers, and uncountable houses, which were left completely destroyed and uninhabitable (Anno Histórico, Tomo III. fl.546, apud Sousa, 1915).
The 1755 earthquake was even more destructive than the previous ones, being considered the greatest earthquake of the entire Christian era to have occurred in the Algarve (COSTA et al. 2005). This earthquake 1 The questions formulated are the following: 5th Did you notice what happened to the sea, to fountains and to rivers ?; 6th Did the sea rise or fall first, and how many "meters". 2 The formulation of question 26 of part I is: If the parish suffered damages caused by the 1755 earthquake, what where they and if they have been repaired? The question 27 of part I of the survey suggests that it should be mentioned everything important not contemplated in this questionnaire (27th And everything else worthy of memory, not contemplated in this questionnaire.). This question was used by some priests to refer the tsunami as the priest of Budens (Vila do Bispo) did. caused a high death toll. The probable number of more than 1,000 deceased in all of the Algarve to an estimated population of 80,000 inhabitants clearly reveals the violence of the earthquake and subsequent tsunami which occurred in this region (COSTA et al, 2005). Nevertheless, in some local communities the proportion of deceased was much higher, for instance in Albufeira and Boliqueime, affecting, respectively, 10% and 7% of the inhabitants (SOUSA, 1919).
In the latter the majority of the deaths resulted from the earthquake, unlike Albufeira where the majority of the population died as a consequence of the tsunami. Here, few were the houses that were left standing and those which resisted were uninhabitable; the sea flooded the land by the outskirts of the village and washed out the Bairro de Santa Ana, which had seven streets, and many other houses without sign of the water flux and reflux or of the houses location (Dicionário Geográfico, 1758).
In other areas the death toll was not as high but the destruction was nonetheless enormous. For example in Vila do Bispo, although only 13 people died (Relaçam, 1755) the village was devastated by the 1755 earthquake, since all houses, except for one, were in ruins. The homeless population had to live outdoors in cold weather deprived of the supplies they had stored, lost under the ruins of the same earthquake (Dicionário Geográfico, 1758). Also all the houses of the village of Aljezur were in ruins, the taller ones completely destroyed, as were the castle and the main church (LOPES, 1841).
The city of Lagos was another of the most affected. In the parish of São Sebastião all the houses of the city were in ruins and in the entire parish only a few onestorey houses were left standing. In this parish 95 people perished, some under the ruins, others taken by the sea. Furthermore in the parish of Santa Maria more than 100 people died, not to mention the people who were spending the day away from the city, which were many, given that this was a holiday (Dicionário Geográfico, 1758).
After 1755 other earthquakes have been felt in this region, but luckily with much lower magnitude. A recent earthquake that hit the Algarve in the morning of 12th of February 2007 measuring 5.8 on the Richter scale alerted to the risk that the occurrence of earthquakes represents to this region and reminded the importance of disaster preparedness.

The 1755 tsunami in the Algarve
Some minutes after the earthquake 3 , the coast of the Algarve was swallowed by a tsunami, which caused a great number of deaths and vast destruction. Along the coast several fishermen were fishing, despite the holiness of the day; but those unfortunate fishermen were all killed by the powerful waves (CASTRO, 1786).
The 1755 tsunami was not an unprecedented event. In fact, there are references to tsunamis caused by other earthquakes which hit the Algarve. A few years earlier, in December 27th, 1722, there was also a tsunami although its effects were only felt locally (SOUSA, 1915).The historical documents testify to some perplexity and lack of knowledge in people's reaction to the 1755 tsunami which contributed to a greater number of deaths in the coastal communities, as it happened for instance with the great majority of the inhabitants of the village of Albufeira, located on the top of a rock, who went down to the beach to find shelter believing the area to be safe. The water hit the beach and swallowed them all (CASTRO, 1786).
But the intensity of 1755 tsunami was enormous. The sea receded in parts more than 36 meters, exposing the beaches; and hitting land with such impetus, ran up inland over 5km, topping the highest rocks; it then receded again and hitting land for three times in a matter of minutes, dragging in the flux and reflux enormous masses of cliffs and buildings; and devastating almost all the coastal populations (LOPES, 1841). Although the effects of the tsunami were felt across the coast of the Algarve, it was particularly destructive in the western part of the region. In fact from Quarteira to Lagos its consequences were dreadful; with great number of fatalities and damage of properties. It swallowed Armação de Pêra, the outskirts of Vila Nova de Portimão and all the low-lying neighbourhoods of the city of Lagos (Relaçam, 1756). In Quarteira, a fishing community, 87 adults and children drowned (OLIVEIRA, 1905). The sea rose up so many meters over its surface that flooded many fields, and when it receded it took a part of the fortress and the entire village of Albufeira, leaving in the wood a large amount of fish (MENDONÇA, 1758).
Lopes (1841) describes that in Lagos the sea rose up to 11 meters nearly topping the city's wall; and washed away all the houses between the coast and the city's wall. It advanced inland over 2.5 km taking with it 5 boats almost to the same distance. The priest of Santa Maria parish reported that the sea destroyed part of the fortress wall and rose more than 7 meters high (ROCHA, 1991). In Armação de Pêra the sea also advanced over 2.5 km, flooding everything (LOPES, 1841).
In the village of Sagres the sea rose up more than 6.6 meters on the beach and in the cliffs around the beach there are records of waves 26.4 meters high, and run up inland over 2.5 km (SOUSA, 1919). The sea touched the fortress of Beliche located at 66meters (Lopes, 1841). In other locations the influence of the tsunami advanced up the rivers and was felt at distances of more than 2.5 km with an estimated height between 11 and 13 meters (LOPES, 1841).
On the eastern coast the effects of the tsunami were not so destructive. In the river of Tavira the water spread apart, in such a way that a caravel that was sailing in the river was left on dry land for a long time allowing the crew to leave the boat on foot (MENDONÇA, 1758). The waves washed away the huts in the beach of Monte Gordo, in which fish was traded also in Conceição de Tavira and swallowed all the islands along the coast up to the beach of Quarteira opening awful mouths, but preserving mercifully the locality of Olhão and the city of Faro (Relaçam, 1756). The little damage suffered by Faro and Olhão (Relaçam, 1756) is not entirely explained by a greater distance from the epicentre. Here the state of the tide and geomorphological configuration of the coastline were crucial factors. The sand islands located right in front of these towns are supposed to have lessened the impact of the wave on those areas. In reality, the capacity of a tsunami to flood and destroy depends largely upon the amplitude (wave height) and wavelength where they are generated but these parameters are also governed by the interaction with sea-floors and the morphology of the coastal strip the moment the tsunami hits (COSTA et al., 2005).
The elevated intensity of the tsunami is also confirmed by the geological records in several locations in the Algarve. One of the most important is Boca do Rio which allows us to conclude that the 1755 tsunami was both erosional and depositional and did not transport significant amounts of sediment from the offshore zone, but rather redeposited material from beaches and dunes, soils and estuarine silts and clays (CHESTER, 2001).
Although the return period of an event with the intensity of the 1755 earthquake is very large, we should point out that the tide gauge of Lagos recorded already in the 20th century two tsunamis: one on the 25th November 1941 measuring I and another on 28th February 1969 measuring II, both of which in the Algarve.
In 1755 run-up heights referred in historical accounts were variable at similar distances from the epicentre. No doubt some of this variance reflects imprecision in the historical data but unlike the timing of onset, run-up height is a reasonably well constrained parameter in the historical record (MENDES et al., 1999). The configuration of the coast, state of the tide, and offshore bathymetry are also crucial factors.

The case-study of Portimão: historical data analysis
The municipality of Portimão comprises the freguesias (parishes) of Alvor, Portimão and Mexilhoeira Grande. The earthquake reached the intensity VIII Mercalli Intensity Scale in Mexilhoeira Grande and X in Alvor and Portimão (SOUSA, 1919, CHESTER, 2001. According to historical references in Alvor the sea inundated 660 meters inland, nearly reaching houses at 30 meters high. The chapel of Nossa Senhora da Ajuda, located on the beach by the harbour, was completely destroyed leaving no traces of its foundations (LOPES, 1841). The sea washed away fishermen who were pulling their nets. (SOUSA, 1919).
The village of Portimão was one of the localities of the Algarve which suffered the most with the 1755 earthquake (Dicionário Geográfico, 1758) and it was also very much affected by the tsunami. In Portimão 6 people died under the ruins and 40 people drowned, 15 of which were female and 19 children (Livro de Óbitos de Portimão apud Sousa, 1919).
In the freguesia of Portimão the sea flowed inland, exceeding the natural barriers, in some areas more than 880 meters, devastating the salt ponds of this village. Since then no salt has been produced there (Dicionário Geográfico, 1758). The harbour of Portimão forms a great mouth berthed between two large rocks, in front of which are the fortresses of Santa Catarina and São João. The first of the two was severely damaged (Dicionário Geográfico, 1758) and the latest was completely washed out (LOPES, 1841).
Through the village of Portimão flowed successively amazing waves, which advanced over 5km up the river. As the waves advanced everything was destroyed. Several boats were carried inland to such great distance that it was not possible to bring them back. Flooding occurred on the outskirts of the village, destroying all the houses and drowning many people, which had sought refuge in the river banks (CASTRO, 1786).
In the parish of Mexilhoeira the damages were not as important because the parish has no direct contact with the sea. The damages caused by the tsunami resulted from the advance of the waves up the River Alvor and other smaller water courses, but nothing worse mentioning happened (Dicionário Geográfico, 1758).
From the accounts about the height of the wave we verify that near the coast it must have got to 20 meters high. It is said that the water almost reached the village of Alvor which is located between 20 to 30 meters above sea level, reaching the Fortress of Santa Catarina (20 meters). The city of Portimão is protected by the rocks of Praia da Rocha, so it is probable that the wave did not reach the height of 20 meters. The Igreja da Misericórdia is the main source for the determination of the wave height ( Figure 2).
As previously referred this church is located 10 meters above sea and the water rose 2.64 meters inside the building. Another building, the Convento de São Francisco located, approximately, 7 metres above the sea, was completely destroyed. These references were  integrated in the map over the Digital Terrain Model (DTM). According to historical sources the height of 12 meters was considered a reliable height for most of the wave but it is possible that in small areas the wave could have reached higher heights as it happened in Fortress of Santa Catarina located on a cliff edge.
Based in the DTM, the lands with altitude below 12 meters and near the coast had been identified, eliminating isolated areas surrounded by lands higher than 12 meters, producing a new DTM of the wave height (inverted) and in a polygon the flooded area ( Figure 3).

The impacts of the 1755 tsunami nowadays
The flooded area should have amounted to 23.75 Km 2 , 13% of the total surface of the municipality. An extension explained by the fact that this is a sandy coast with a reduced number of cliffs and by the existence of two rivers, natural boundaries of the municipality, serving as hallways allowing the advance of the waves.
Since 1755, two key factors have increased the vulnerability in Portimão municipality. Seaside settlement is no longer solely based on livelihood but on urbanization and tourism associated with the attraction of enjoying living and holidaying literally in front of the sea.
Nowadays, the potentially flooded area would put at risk, approximately, 21,845 inhabitants, according to 2001 Census data. The advance of a 12 meters wave up River Arade would affect without a doubt the low area of Portimão, where the most densely populated areas (more than 300 inhabitants/ ha) are located ( Figure 4). Here the population affected would amount to approximately 18,303 inhabitants. In the others parishes less inhabitants would be in danger: 2,403 inhabitants in Alvor and 1,139 inhabitants in Mexilhoeira.
Since Portimão is an area where tourism is the most important economic activity, we can conclude that the population at risk would actually be much higher. Although there is no statistical record of the temporary population and it varies tremendously during the year, if we take into consideration the water and electricity consumption in certain seasons, it is possible to say that the population doubles in some periods.
We estimate that approximately 5,000 buildings would be affected, many of which would be destroyed or severely damaged, including hotels. However, the vulnerability of the hotel units is highly variable and the units located in the low-lying parts of the city of Portimão would be the most seriously affected. We identified 67 hotels, of which 26 are located in the potentially flooded area ( Figure 5).The newest resorts are located in highly susceptible areas, as it is the case of the hotel, restaurants and marina built in the mouth of the River Arade, which would represent the first structure to be struck by the waves of a tsunami (Photo 1).
In a disaster situation, the management of rescue means and security forces is essential. From our analysis, it is possible to conclude that in the established scenario the structures of civil protection agents would not be severely damaged ( Figure 6). Only the police headquarter is located in a potential flooded area but fire-fighter brigades and other security forces will not be affected.
When considering the health facilities, the primary care health centres would be affected but not the hospitals. Also located in the potential flooded area are the prison, and the local building of Red Cross.
The aerodrome, important equipment allowing easy access of rescue means or the evacuation of inhabitants, is located in freguesia de Alvor in an area easily affected, since it is located at a height of 5 metres and on the bank of the Alvor River, up which the tsunami would advance.
In addition, the road network itself would be severely affected and significant stretches of essential regional routes would either be flooded or destroyed. The determination of the routes potencially affected and the routes possible to be used in case of accident is vital both for the evacuation and rescue of the populations affected. We can thus conclude that faced with a tsunami similar to the 1755 event, the region would be considerably dependent on external rescue means.
These analyses help us to demonstrate the population's exposure to risk and the urgent need to develop protection mechanisms, including a greater attention to land management efforts, especially as far as the location of priority equipment is concerned, but also the establishment of a building code, because if there have been tsunamis in the Algarve in the past, another tsunami could occur in the region in the future. They are a summarised description of the virtuosity of our approach to emergency planning, although it is also possible to include other variables such as, water and electricity infra-structures. Also we could necessarily improve the quantitative approach.

Conclusion
In the Southwest part of Algarve the waves were higher than in Lisbon and the flooded area was much larger. Although the average return period of a tsunami similar to the one that hit the Algarve in 1755 is very great, probably greater than 1000 years, it may occur at any time provided that the seismic source is totally or partially located in the sea and that the magnitude measures more than 7.5 (COSTA et al., 2005). In that case a tsunami would rapidly arrive, meaning that a quick perception and reaction would be necessary to minimise the damages. In the 1755 event it is  Photo 1 -The vulnerability of these buildings is very high.
supposed that the tsunami arrival time was 7±16 minutes (BAPTISTA et al., , 2005. Thus, the delimitation of the flooded area is an important tool for emergency planning. It is obvious that the simulation we present here is valid for an earthquake with the same magnitude and tectonic source of the 1755 event. Of course, other earthquakes with offshore epicentre but different magnitude and location will lead to distinct flood heights and inundated areas. In fact, the 28th February 1969 earthquake with a magnitude of 7.5 (MARTINS and MENDES-VITOR., 2001) or 7.9  and epicentre at the Horseshoe Abyssal plain , south of Gorringe Bank (FUKAO, 1973) generated a tsunami with weak intensity that did not put at risk the coastal communities in the Algarve.
Although we have not used modulation to assess the energy of the tsunami, the simple identification of the flooded area is fundamental because the more extensive the area affected is, the higher the height of the wave and the energy transported are (COSTA et al., 2005).
Also we must remember that run-up heights are not just a function of distance from the epicentre. In 1755 run-up heights were variable at similar distances from the epicentre. No doubt some of this variance reflects imprecision in the historical data but unlike the timing of onset, run-up height is a reasonably well constrained parameter in the historical record (MENDES et al., 1999). The configuration of the coast, state of the tide, and offshore bathymetry are also crucial factors.
We assess the potential impacts of a wave 12 meters high but it is also possible to define other scenarios with obviously different impacts. The delimitation of flooded area besides serving as basis to the definition of scenarios and the determination of emergency planning tasks, may also serve as a guiding principle for land management.
Actually, despite the low rate of occurrence of disastrous tsunamis, their impact is so great and their consequences so dramatic that their existence should be taken into account in assessing natural risks in coastal areas (BAPTISTA et al., 2005).

Introduction
The Serra do Marão (as it is called locally) is a mountainous massif situated in the North of Portugal formed mainly of Palaeozoic structures. It is strongly marked by the Hercynian orogeny, which is responsible for a large number of faults and fractures. Neo-tectonic phenomena, which are quite common in northern Portugal, influenced this structure and created new faults which made the morphostructure of this area even more complex.
The Marão is part of a mountainous range with a north-south direction and is therefore parallel to the Atlantic coast, which explains the heavy rainfall on the slopes turned to the west in contrast to a much lower precipitation on the slopes facing the east.
During the Quaternary the action of the cold weather was very important in terms of the morphogenetic processes that acted on the slopes. On the one hand, the deposits regularised the slopes; on the other, they conditioned the present dynamics, as will be demonstrated in this paper.

Geomorphologic features
The Serra do Marão is one of the most impressive topographic structures that can be singled out in the vast orographic range that separates the northwest of Portugal from the plateaux of Trás-os-Montes (Fig.1). It is one of a series of mountain ranges that are traditionally seen as a whole: the mountainous range that separates Minho from Trás-os-Montes, the two most characteristic regions in northern Portugal. Starting from the frontier of Galicia and moving southwards several mountain ranges are to be found: Peneda (1373 m), Gerês (1300 m), Larouco (1527 m), Cabreira (1262 m), Alvão (1285 m) and Marão (1416 m). The latter borders on the Douro valley. Still another mountain range can be added to the list -Montemuro (1381 m) but it lies south of the river Douro. The highest points in northern Portugal are to be found in this range.
This group of mountain ranges and elevated plateaux constitute (not only on account of their height, but also of their bulk) an obstacle which determines a more or less rapid transformation in the Atlantic character of the landscapes; because of that it is commonly termed condensation barrier. The Marão is situated in one of the areas of the country with the heaviest rainfall and is simultaneously the one that offers the strongest contrasts in terms of the distribution of temperature. These two characteristics -heavy rainfall and large temperature range -make it unique among the different/various Portuguese mountain ranges.
The action of tectonics is fundamental to the understanding of the formation of the Marão. The Hercynian orogeny was the main factor responsible for the uplift and folding of the metasediments of the Pre-Ordovician, the Ordovician and the Silurian (Fig. 2). It was also during this orogeny that the intrusion of the different granitic rocks, which are found on the periphery of the mountain range, took place. These orogenic movements explain many of the fractures and faults that are still today important features in the morphology of the area.
In spite of the importance of the Hercynian movements in this mountain range, there is no doubt that the present-day relief forms are related with the alpine orogeny. This may have had a basically epeirogenetic action which contributed to the uplift both of the mountain ranges of the Marão and the Alvão. Consequently, these two mountain ranges function as a horst both in relation to the group of tectonic depressions where the river Corgo flows and also to the graben of the river Tâmega. In fact, the effects of the orogeny are still to be felt and the neo-tectonics goes on acting on the present relief forms (PEDROSA, A., 1993(PEDROSA, A., , 1994. That is, tectonic action continues to transform them; thus it is a permanent risk that must be taken into account, notably when building infrastructures such as roads, railways, dams and even houses.
Lithology is another important science that can explain some relief forms that are found in the Marão. The different granitic rocks, for example, are responsible for the great number of alveoli that exist there. No matter what their form and size are or the height at which they may be found, they are undoubtedly related to the different types of granite. They also account for a series of microforms such as tors, block chaos, "vascas", "caneluras" which are clearly connected with the physical, chemical and mineralogical characteristics of these rocks.
The quartzite is a type of rock that has a special importance in the morphology of the Marão (PEDROSA, 1993) in spite of being quite often found in the upper parts of its slopes. They are almost always found halfway up the slopes, which can be explained by the action of tectonics. Nevertheless, although this type of quartzite outcropping is more exposed to erosive processes due to its situation, it undoubtedly contributes to the fact that though the inclination of the slopes is always steep the erosion acts more slowly.
In the schist areas there are several depressions of tectonic origin, but the fundamental characteristic is in this case the regulation of the slopes by deposits related to morphogenetic processes which took place in the recent Quaternary and which were caused by the freeze/thaw action. As such, they correspond to colder climates, which are responsible for a very rapid evolution of the slopes (Photo1).

Origin of the deposits
The effects of the different morphogenetic processes which happened in the past can only be inferred from vestiges that are still observable nowadays: the slope deposits. It is not always easy to interpret them. Many have been destroyed, thus proving the existence of a powerful morphogenetic action which is related to colder climates in the course of the Quaternary.
The deposits which were found are important evidence of the evolution of the slopes, not only at a high altitude (PEDROSA, A.,1994a) but also in areas at low altitude in places near the coast (PEDROSA, A.,1989(PEDROSA, A., , 1994cREBELO, F. & PEDROSA, A.,1993). The climate must then have been much more unstable than it is now with temperatures that made gelifraction in situ possible, above all in higher places where vegetation was scarce or non-existent.
The material must have been made transported along the slopes by gelifluction associated with a cold, dry climate (DAVEAU, S., 1973;REBELO, F. 1986;CORDEIRO, A. R., 1986a;PEDROSA, A., 1993PEDROSA, A., , 1994a; it is possible to observe the existence of coiffes (VALADAS, B., 1984). Nevertheless, the above mentioned movement must not have been carried out only by the action of ice but also by means of more or less generalised solifluctions or even by debris flows and sheet floods. The existence of palaeogullies may lead to the conclusion that there were larger quantities of water, at least during part of the year; this would make possible the formation of gullies in the materials accumulated at the foot of the slopes by other morphogenetic processes.
Thus, the accumulation of materials along the slopes would have been necessary for the formation of such palaeogullies but the morphogenetic processes could have been different. Absolutely necessary for their formation was, however, the alternation of drier, colder periods with others which were damper and hypothetically less cold (PEDROSA, A.,1993).
The palaeosols found seem to indicate climatic fluctuations during which the existence of colder climates, hypothetically drier, contrasted with periods of higher biostasis which might mean that there were higher quantities of rainfall and/or milder temperatures.
The oldest deposit found in the Marão is prior to 28000 BP; it was possible to find a palaeosol that fossilises it dating back to 28440 +/-490 BP. It may correspond to the so called Würm II (PEDROSA, A.,1993). It is a deposit with heterometric characteristics, probably the result of the deposition of material transported by debris flow, although the possibility of its having resulted from glacial action is not to be put aside (Fig. 3).
After the deposition of the materials that form that deposit, the climate may have evolved; it may have become moister and the temperatures milder. This may have been responsible for the pedogenesis and the appearance of the above mentioned paleosoil. The climate must have become slightly warmer in comparison with the prior phase; this fact gave way to the pedogenesis and the colonisation of the slopes by vegetation, making them reach a dynamic equilibrium.
Later on the climate must have turned cooler, which led to the formation of new slope deposits which fossilised the palaeosoil. Due to the disappearance of the vegetation, the slopes must have been more exposed to the action of fragmentation by the action of ice. The analysis of the deposits allows the observer to conclude that the morphogenetic processes must have been not only mass movements of the solifluction type but also debris flows. And again comes a period of pedogenesis which dates back to around 21340 +/-350 BP.
The next phase corresponds to the influence of the "Upper Würmian Full Glaciation" (Pleniglaciar superior Würmiano), the last glacial stage of the Pleistocene. This phase was extremely important in the evolution of the slopes in northern Portugal. In fact, there is a large number of slope deposits whose origin is to be found in this period (Photo 2). The climate must have been cold and relatively dry because it is responsible for the gelifraction of the rocks (DAVEAU, S., 1973;REBELO, F., 1986;CORDEIRO, A. R.,1986b;REBELO, F., & PEDROSA, A., 1993).
The analysis of these deposits demonstrated that the main process of transportation along the slopes was the action of the ice -gelifluxion, further proved by the existence of coiffes (VALADAS, 1984). Other processes such as the action of gravity are, however, not to be ignored. The study of the stratified slope deposits does not exclude the possibility of other movements such as debris flows at least in some parts of the slopes. With the beginning of the Tardiglacial (CORDEIRO, A. R.,1990) the climate must have become warmer and the rainfall heavier, which must have altered the morphogenetic processes; now the most common are those connected with the runoff and fluvial dynamics instead of those related to the action of the ice.
Parallel to the significant movement of materials which occurred in this period there was an extensive movement of materials resulting from prior morphogenetic processes transported more by fluvial action than by the action of the ice and solifluctions in periods of thawing. (PEDROSA, A.,1993). They contributed to the filling of the bottoms of small valleys and the regularisation of a large part of the slopes (REBELO, F., 1975;REBELO, F. & PEDROSA, A., 1993;PEDROSA, A., 1988PEDROSA, A., , 1993. However, the climate warming verified at the beginning of the Tardiglacial cannot have been linear (COUDE-GAUSSEN, G.,1981;FERREIRA, A. B., 1985;Rebelo, F., 1985Rebelo, F., , 1986CUNHA, L., 1988;CORDEIRO, A. R., 1986aCORDEIRO, A. R., , 1988.Today the existence of at least three moments in the last Tardiglacial are recognised (NONN, 1966;GUILLEN, 1962;DAVEAU, 1973DAVEAU, , 1978, with an interphase and a recurrence of a very cold, dry phase (JORDA, 1980;GARMENDIA, 1989;VLIET-LANOE, 1988). In fact, it seems that the Tardiglacial is characterised by the existence of several climatic fluctuations (CORDEIRO, 1990;Pedrosa, A., 1994c); MARTINS, 1999), confirmed by the features of the formation of Covelo do Monte. About 11000 BP a new climatic crisis must have occurred (CORDEIRO, A. R. , 1990;GULLIEN et al., 1978;GARMENDIA, 1984); it would have led not only to the occurrence of processes related to the action of the ice but also to generalised solifluctions which may have caused the remixing of deposits formed by prior processes.
It is possible to find a great heterogeneity in the Tardiglacial deposits, which depends on the exposure of the slopes, on their rock substratum and on their inclination (PEDROSA, A., 1994b;MARTINS, 1999); but the fundamental factor must have been the snowfall during a part of the year and its thawing in the warmer period, which was probably the time of the heaviest rainfall. The most important processes must have been the solifluctions which were responsible for the regularisation of a great part of the slopes as well as the occurrence of slides especially extensive debris flows, whose marks can still be traced on the bottoms of the valleys (Photo 3). However, it is not possible to find mature soils; the majority are entisols and are quickly affected by small climatic changes. A new warming seems to have reached its peak by 8000 BP as it was possible to find in the Marão some palaeosols which are contemporary with this period known as Holocene climatic optimum.
Around 5000 BP when the sub-Boreal period began (GOUDIE, 1979) an increase in erosion must have taken place; this increase is probably connected with a drop in temperature as well as with some human activities, especially the burning of land to renew the vegetation and obtain new pastureland, which would increase the power of such erosion processes as runoff and splash (PEDROSA, A., 1993). Solifluctions must also have occurred in some parts of the slopes, above all in higher places facing north. In the Marão the best example of this period is the formation of Lameira.
Gravity slumping associated with a massive congelifraction of rocks may correspond to different generations of these forms or, at least, to the existence of different periods which activated their formation and evolution. In the Marão they are generally found at heights superior to 700m and are sometimes associated with steep inclines, especially on shady slopes. An important phase in the origin of these forms may have been the short glacial age, which occurred in the 17th and 18th centuries (LADURIE, 1983). This process is still active nowadays in places situated on shady slopes where the bare rock appears on the surface without any soil or vegetation covering it (PEDROSA, A.,1993).
The importance of slope deposits for the present slope morphodynamics The geomorphologic dynamics of the mountain ranges of northern Portugal is very complex since both the factors and the intervening agents as well as their interaction are diverse. Even if only the natural agents are taken into account, there is still the interaction of several factors whose relative importance varies from region to region or even from place to place, making the active morphogenetic processes different. It is indispensable to understand the behaviour of the fundamental climatic elements -precipitation and temperature -to explain the action of the processes which are responsible for the present evolution of the slopes, but it is also necessary to take into account factors such as the lithology, the morphology, the biogeography and the surface formations.

The importance of the formation of gullies
In these areas the formation of gullies is clearly conditioned by the existence of a slope deposit, which works as a regulator because there are no trees or quite often because man has acted on this terrain.
The examples that will now be given show clearly what has just been said.
In July 1992, on a slope near the village of Paradela do Monte (municipality of Santa Marta de Penaguião) there was a very heavy shower. Unfortunately, it is not recorded, owing to the small geographic extension of the area on which it rained and the non-existence of pluviometres, but it is estimated to have reached values very nearly 30 mm while it lasted (about 1 hour); in consequence, two big gullies appeared (Photo 4). slope (30º), the growth of new vegetation is rather slow. The whole slope is regularised by a deposit which in some places is more than 2 meters thick. More or less in the middle of the slope two tracks were opened on the slopes to lay a system of waterpipes. This was the background in which the slope gullying began. At first the runoff flowed along one of the tracks and formed gullies. Later the drainage stopped flowing along the tracks, turned in the direction of the steeper incline and formed two gullies whose depth is clearly conditioned by the depth of the deposit which regularises the slope. As soon as the water reached the schists in situ the gully thalweg could not become deeper; it became wider and the water transported the slope material which was not well consolidated. One of them deepened a pre-existing gully while the other developed at first parallel to a small valley that existed on the slope, later using only its final part; as a consequence, its thalweg became deeper.
In order to calculate the quantity of material transported, the length, the width and depth of the two gullies, they were measured in different sections. The results of these measurements were as follows: in one of the gullies the volume of material transported was about 465 m 3 while in the other it was a bit more: about 570 m 3 (Chart 1). All this material must have been transported only in the short period of time during which the rain poured down; one of the gullies has a cone of dejection just above the road (Photo 5), formed by the coarser materials, while the material transported by the other crossed the road and ran down the slope into the main brook. In July 1991 there had been a fire which had completely destroyed the vegetation covering that slope. Due to the altitude and the inclination of the slope (30º), the growth of new vegetation is rather slow. The whole slope is regularised by a deposit which in some places is more than 2 meters thick. More or less in the middle of the slope two tracks were opened on the slopes to lay a system of waterpipes. This was the background in which the slope gullying began. At first the runoff flowed along one of the tracks and formed gullies. Later the drainage stopped flowing along the tracks, turned in the direction of the steeper incline and formed two gullies whose depth is clearly conditioned by the depth of the deposit which regularises the slope. As soon as the water reached the schists in situ the gully thalweg could not become deeper; it became wider and the water transported the slope material which was not well consolidated. One of them deepened a pre-existing gully while the other developed at first parallel to a small valley that existed on the slope, later using only its final part; as a consequence, its thalweg became deeper.
In July 1991 there had been a fire which had completely destroyed the vegetation covering that slope. Due to the altitude and the inclination of the Chart 1-Dimension and Volume of material transported by the gullies.
Photo 5 -Detail of the material deposited in the slope after gullying.

Width Depth
The importance of mass movements Mass movements can have a greater or smaller complexity because they can be the result of different processes which can only be identified by an analysis of each individual case. Roughly speaking, they can go from the moving of small individual particles to the fall of big stones and slow or fast flows.
However, these movements cannot be explained only by adverse meteorological conditions, as neither heavy rainfalls nor all the long periods of precipitation always give rise to soil movements. Therefore, it is necessary to look for other factors which acting together make it possible to explain the onset of the initial movement which causes the earth-flow .
In several recent studies different authors (REBELO, 2001;PEDROSA, 1993PEDROSA, , 1994aPEDROSA, , 1994bPEDROSA, LOURENÇO, FELGUEIRAS, 2001;PEDROSA, MARTINS, 2001;PEDROSA, BATEIRA, SOARES,1995;PEDROSA, MARQUES, 1994) have tried to systematise the factors that are of greater importance for the onset of these processes in northern Portugal. Natural factors that can influence the occurrence of debris flows and mass movements have been considered up to this moment. Nevertheless, factors related to human interference are more and more frequent. Some are connected with the subdividing of estates into plots for the building of houses, others with the building of roads and railways which imply extensive back filling (aterro) and excavation (desaterro) of land; these alter the conditions of equilibrium on the slope and are responsible for the occurrence or, at least, for starting the initial process.
Humankind has been a geomorphological agent for a long time and in this role is becoming more and more important. Human influence creates "humanised landscapes" which do not always safeguard the operation of the natural processes and thus contribute to situations which, left to nature, might eventually happen but much, much later.
It is unquestionable that many of these movements (not only the individualised ones such as landslides, earth falls, mudflows, and so on, but also those that act together) almost always happen after periods of heavy rainfall which leave the soil, the alteration mantles (mantos de alteração) and the slope deposits saturated or near saturation, so creating conditions to initiate the earth flow.
Besides, the infiltration resulting from the rainwater activates the underground circulation which sometimes respects former palaeotopographies, i. e., former small river valleys fossilised by slope deposits. The water manages to trickle between the rock, little or not altered, and the overlying deposits. When these underground flows are strong and are associated with the water infiltrated directly from the rainfall, they can contribute to the fluidity of the overlying deposits and, in doing so, to create conditions for their starting a movement which often extends downstream through flows containing a higher or smaller quantity of mud.
Besides the hydroclimatic conditions, structural aspects both lithological and tectonic are sometimes fundamental factors, often playing an important part in that they condition the type of movement, its size and the evolution of the area affected.
Quite often it is the net of faults and fractures that contributes to facilitate the alteration of the rocks and allows a more efficient infiltration of water which reaches progressively deeper levels and ends up by making the surface mass movements easier.
The geomorphological factors are the most varied and those that have more significant repercussions when connected with other factors -natural and human.
The steepness of the slope is another morphological factor of great importance for the onset of mass movements. The above mentioned authors speak of steep inclines almost always superior to 30% as being one of the morphological factors decisive for the fast evolution of the slopes and it is one of the criteria that may provide important elements for the definition of areas of potential risk.
Breaks in the slope, above all when they contribute to the increase of the incline downstream, are a local factor responsible for the worsening of the conditions of the occurrence of movements; because of that, they deserve a special reference.
Besides the incline, the form of the slope also plays an important part in the creation of conditions favourable to the development of processes of slope evolution.
Slope regulation by surface formations is undoubtedly one of the factors that may contribute in a more decisive way to the development of these mass movements, which according to their characteristics can lead to movements sometimes slow sometimes fast. Thus, when the surface formation corresponds to deep alteration mantles, capable of absorbing large quantities of water, it can lead to extensive mass movements. In fact, the great thickness of the alteration mantle or of the deposit favours the water infiltration; the saturation point can then be reached and the soil may begin to collapse. The great absorption power may slow down the beginning of the movement and bring about its development in stages over the course of several days.
Another frequent situation has to do with the thin slope deposits of the Quaternary, rich in argillaceous material with a great capacity to absorb water, lying on rocks which are little altered and which quite often act as an efficient sliding surface. In the schist and quartzite areas of the mountain ranges in northern Portugal, this is one of the most efficient natural processes of slope evolution. In fact, when heavy rain falls for long periods of time on steep slopes regularised by slope deposits, mass movements such as debris flows happen frequently.

Debris flows in the Serra do Marão
Debris flows are a frequent occurrence on the slopes of the Marão. They are related to the quantity of rainfall, the steep inclination of the slopes and above all to the regularisation of the schist slopes by the deposits of the Quaternary which have already been mentioned in the first part of this paper. In fact, it is on the discontinuity surface which coincides with the contact between the slope deposits and the bedrock that the failure occurs and the debris flow begins.
In fact, surface formations constituted by the slope deposits of the Quaternary allow the infiltration of the rainwater that then flows through the deposit materials but above all in the surface of contact between them and the metasedimentary rocks, which owing to their impermeability, lead to the accumulation and flow of the infiltrated water. Thus, when the period of rainfall is long, the fluidity of the overlying deposits may increase, causing the rupture.
In 1992 several debris flows occurred in consequence of a long period of precipitation (A. Pedrosa, 1993); they happened near the villages of Montes e Póvoa da Serra, whose slopes are steep and covered with trees; but the fundamental reason was the fact that they were regularised by deposits (Photos 6 and 7).
Later, in 1997, after a few very rainy days, some more debris flows occurred in slopes with the same characteristics as those already described. This is the case of Portal da Freita (Photo 8) and of Carvalhada de Baixo (Photo 9) both near the village of Montes.
Again in 2001, after a long period of rainfall, there were debris flows down several slopes; all of them had the same characteristics: a steep inclination and the presence of a surface formation with a thickness of 1.5 / 2 metres which regularises the slopes. In that same year, besides the occurrence of debris flows in new places, there were also new movements in the locations of previous flows, such as the ones in Póvoa da Serra or in Montes.
Photo 6 -Characteristics of the debris flow of Póvoa da Serra.
Photo 7 -Another detail of the debris flow in Póvoa da Serra.
Photo 9 -Debris flow in Carvalhal de Baixo where it is possible to observe that the place of the flow began coincides with slope are regularised by deposit.
Photo 8 -Detail of the place where the debris flow of Portal da Freita began, whose failure is clearly related with the contact of the deposit with the bedrock.
The one that had the most serious consequences happened place near Mesão Frio (Volta Grande) (Photo 10). It reached road no N101 (Amarante -Peso da Régua), dragging two cars that were driving along it and causing the death of one of the drivers. This example shows clearly the risk of building roads on the foot of slopes with the characteristics that have been described because processes of rapid change in the characteristics of the slopes often occur in these circumstances, leading to a greater instability of the regolith materials that regularise the slope. one goes from steeper slopes to gentler ones, one may observe the deceleration of the erosion processes in favour of processes of accumulation of materials (YOUNG, 1969).
Slope inclinations have great influence on the dynamics of any present-day morphogenetic process; therefore, it is an ever present factor in the analysis and understanding of the morphogenetic processes and their action.
On the slopes on which these flows occurred the mean inclination varies between 30º and 40º which is an important element in the onset of the whole process. In some places the gradient is even higher: for example, near Póvoa da Serra the mean slope inclination is over 40º. However, near the scars in the place where the debris flow began the angle is over 60º.
But on the slopes of the Marão it is the fact that they are regularised by slope deposits that are very different in size that facilitates the infiltration of rainwater even on steep slopes. The water reaches quickly the plane of contact with the bedrock, which is relatively impermeable. That is where the rupture occurs which allows the debris flow to begin.
The study of the present morphogenetic processes may contribute to knowledge about the natural hazards (MARTIN, 1987;REBELO, 1991bREBELO, , 1994BRYANT, 1991). Knowing the elements that interact on the slopes is fundamental; they must be understood from a dynamic point of view, distinguishing between the major and the minor factors.
The geomorphologist must know as deeply as possible the combinations and ways in which these factors act. But although it is possible to map areas where landslides may take place, the forecast of the occurrence of such phenomena is based merely on a probability and not on a precise indication both in time and in space.
However, the study of a growing number of cases related to specific morphogenetic processes makes it possible to get a deeper insight into the way these processes operate, which will enable the authorities to mitigate their harmful consequences, to establish a scale of risks and draw maps showing the vulnerable areas in the landscape. Photo 10 -Location of the occurrence of the debris flow of Volta Grande which affected road N101 and dragged two cars causing one death among the car occupants.

Conclusion
The behaviour of climatic elements is important for the understanding of the morphogenetic processes. The analysis of such elements cannot confine itself to the mean values because it is important to know their variation, not only during one year but also over a larger number of years, especially on account of the consequences that that variation may have on the morphogenetic processes (REBELO, 1983;BRYANT,1991;FAUGÈRES, 1991).
Mass movements do not always coincide with the day of the heaviest rainfall; quite often they occur on days on which the precipitation is scarce or even nonexistent. So it is indispensable to observe the pluviometric values of periods further from the day on which the movement occurred. Although there may be a relationship between the occurrence of very heavy rainfall and the occurrence of movements the prolonging of heavy rain over many days causes soil saturation (FLAGEOLLET, 1989). Thus, in areas with large quantities of argillaceous material the relationship of heavy rainfall to earthflow movement is not so direct, because its water absorption capacity is very high, which retards the occurrence of the slip.
Another important factor in the dynamics of the morphogenetic processes is morphology. Thus, when