Fault tolerance logging-based model for deterministic systems

Fault tolerance allows a system to remain operational to some degree when some of its components fail. One of the most common fault tolerance mechanisms consists on logging the system state periodically, and recovering the system to a consistent state in the event of a failure. This paper describes a general fault tolerance logging-based mechanism, which can be layered over deterministic systems. Our proposal describes how a logging mechanism can recover the underlying system to a consistent state, even if an action or set of actions were interrupted mid-way, due to a server crash. We also propose different methods of storing the logging information, and describe how to deploy a fault tolerant master-slave cluster for information replication. We adapt our model to a previously proposed framework, which provided common relational features, like transactions with atomic, consistent, isolated and durable properties, to NoSQL database management systems.

Assessing the combined effect of land management and wildfire on runoff and soil erosion in north central Portugal

Strong and sometimes extreme responses in runoff and soil erosion following wildfires have been reported worldwide. However, in the case of North-Central Portugal, little research had been carried out regarding the hydrologic and erosive impacts of several land management activities in recently burnt areas (such as ground preparation, post-fire logging or post-fire mitigation treatments). This study aims to assess post-fire runoff and soil erosion response on Eucalypt and Maritime pine plantations during the first, second and third years following wildfires. The effect of several pre-fire ground preparation operations (ploughed down-slope, contour ploughed and inclined terraces), post-fire logging activities (on both the eucalypt and pine plantations), as well as the application of hydromulch (a post-fire emergency treatment) on overland flow and soil erosion were compared to burnt but undisturbed and untreated areas. The intensive monitoring of runoff, soil erosion and selected soil properties served to determine the main factors involved in post-fire runoff and soil erosion and their spatial and temporal variation. Soil water repellency deserved special attention, due to its supposed important role for overland flow generation. Repeated rainfall simulation experiments (RSE’s), micro-scale runoff plots and bounded sediment fences were carried out and/or installed immediately after the wildfire on seven burnt slopes. Micro-scale runoff plots results under natural rainfall conditions were also compared to the RSE’s results, which was useful for assessing the representativeness of the data obtained with artificial rainfall. The results showed comparable runoff coefficient (20-60%) but lower sediment losses (125-1000 g m-2) than prior studies in Portugal, but especially outside Portugal. Lower sediment losses were related with the historic intensive land use in the area. In evaluating these losses, however, the shallowness and stoniness of the soils, as well as the high organic matter fraction of the eroded sediments (50%) must not be overlooked. Sediment limited erosion was measured in all the ploughed sites, probably due to the time since ploughing (several years). The disturbance of the soil surface cover due to post-fire logging and wood extraction substantially increased sediment losses at both the pine and eucalypt sites. Hydromulch effectiveness in reducing the runoff (70%) and sediment losses (83%) was attributed to the protective high coverage provided by hydromulch. The hydromulch significantly affected the soil cover and other soil properties and these changes also reduced the soil erosion risk. The rainfall amount was the main factor explaining the variance in runoff. However, a shift from rainfall amount to rainfall intensity was detected when either the surface cover or the infiltration capacity (hydrophilic conditions) increased. Sediment losses were controlled by rainfall intensity and surface cover. The role of soil water repellency on runoff generation was not consistent; the overall repellency levels alone were not enough to assess its hydrological impact. Soil water repellency explained runoff generation in the specific-sites model better than in the overall model. Additionally, soil moisture content was a better predictor for soil water repellency than antecedent rainfall. The natural rainfall results confirmed that RSE’s were able to capture the specific sediment losses and its organic matter content as well as the differences between the ploughed and unploughed sites. Repeated RSE’s also captured the seasonal variations in runoff and sediment losses attributed to soil water repellency. These results have implications for post-fire soil erosion modelling and soil conservation practices in the region, or areas with the same land use, climate and soil characteristics. The measured sediment loss, as well as the increasing frequency of ploughing in recently burnt and unburnt eucalypt stands, suggests ploughing is not an effective as a soil conservation measure. Logging activities with less impact are recommended in order to maintain the forest litter protecting the soil surface. Due to its high effectiveness in reducing runoff and soil erosion, hydromulch is recommended for highly sensitive and vulnerable areas.

Soil erosion mitigation following forest wildfires

This study aims to measure the effectiveness of four post-fire emergency techniques for reducing overland flow and soil erosion on the central-Portugal typical forest. The selection and development of these techniques was based on the review of the scientific background, but specially after checking throughout field rainfall simulation experiments which factors were the key for runoff and soil erosion on the specific case of high repellent soils. The forest residue mulch, a new treatment never tested before, was highly effective in reducing runoff and soil erosion in recently burnt eucalypt forest. The logging slash mulch had no obvious effect, but it was attributed to the small amounts of runoff and sediments that the untreated plots produced due to the extensive needle cast following a low severity fire. The hydromulch, a mixture of water, organic fibres, seeds, nutrients and a surfactant used in cutted slopes rehabilitation was also highly successful and was specially indicated for especially sensible areas. The utilization of polyacrylamides, a chemical agent with good performance in agricultural erosion, was not successful in post-fire runoff and soil erosion control, once that did not alter the most important key factor for soil erosion: the ground cover. The development of a new fibre optic turbidity sensor was a successful development on the soil erosion determination methodology, and its patent is being processed in the mean time.

Post-fire vegetation regeneration in Portugal : implications for management

This thesis aims at improving the knowledge on the post-fire vegetation regeneration. For that, forests and shrublands were studied, after forest fires and experimental fires. Maritime Pine (Pinus pinaster) recruitment after fire was studied. Fire severity was evidenced as a major effect on this process. High crown fire severity can combust the pines, destroying the seed bank and impeding post fire pine recruitment. However, crown combustion also influences the post-fire conditions on the soil surface, since high crown combustion (HCC) will decrease the postfire needle cast. After low crown combustion (LCC) (scorched rather than torched crowns), a considerable needle cover was observed, along with a higher density of pine seedlings. The overall trends of post-fire recruitment among LCC and HCC areas could be significantly attributed to cover by needles, as well by the estimation of fire severity using the diameters of the burned twigs (TSI). Fire increased the germination from the soil seed bank of a Pinus pinaster forest, and the effects were also related with fire severity. The densities of seedlings of the dominant taxa (genus Erica and Calluna vulgaris) were contrastingly affected in relation to the unburned situation, depending on fire severity, as estimated from the degree of fire-induced crown damage (LCC/HCC), as well as using a severity index based on the diameters of remaining twigs (TSI). Low severity patches had an increase in germination density relatively to the control, while high severity patches suffered a reduction. After an experimental fire in a heathland dominated by Pterospartum tridentatum, Erica australis and E. umbellata, no net differences in seedling emergence were observed, in relation to the pre-fire situation. However, rather than having no effect, the heterogeneity of temperatures caused by fire promoted caused divergent effects over the burned plot in terms of Erica australis germination – a progressive increased was observed in the plots were maximum temperature recorded ranged from 29 to 42.5ºC and decreased in plots with maximum temperature ranging from 51.5 to 74.5ºC. In this heathland, the seed density of two of the main species (E. australis and E. umbellata) was higher under their canopies, but the same was not true for P. tridentatum. The understory regeneration in pine and eucalypt stands, 5 to 6 years post fire, has been strongly associated with post-fire management practices. The effect of forest type was, comparatively, insignificant. Soil tilling, tree harvesting and shrub clearance, were linked to lower soil cover percentages. However, while all these management operations negatively affected the cover of resprouters, seeders were not affected by soil tilling. A strong influence of biogeographic region was identified, suggesting that more vulnerable regions may suffer higher effects of management, even under comparatively lower management pressure than more productive regions. This emphasizes the need to adequate post-fire management techniques to the target regions.