Modelling shallow landslide susceptibility: a new approach in logistic regression by using favourability assessment

A new method for estimating shallow landslide susceptibility by combining Geographical Information System (GIS), nonparametric kernel density estimation and logistic regression is described. Specifically, a logistic regression is applied to predict the spatial distribution by estimating the probability of occurrence of a landslide in a 16 km² area. For this purpose, a GIS is employed to gather the relevant sample information connected with the landslides. The advantages of pre-processing the explanatory variables by nonparametric density estimation (for continuous variables) and a reclassification (for categorical/discrete ones) are discussed. The pre-processing leads to new explanatory variables, namely, some functions which measure the favourability of occurrence of a landslide. The resulting model correctly classifies 98.55% of the inventaried landslides and 89.80% of the landscape surface without instabilities. New data about recent shallow landslides were collected in order to validate the model, and 92.20% of them are also correctly classified. The results support the methodology and the extrapolation of the model to the whole study area (278 km²) in order to obtain susceptibility maps.     

Land use dynamics in Brazilian La Plata Basin and anthropogenic climate change

The La Plata Basin (LPB) is one of the most important regions for agriculture and livestock production in South America, playing a central role in the world food production and food security. Within its borders is also located the whole Brazilian Pantanal region. Identifying the most important land use sectors in LPB as well as the changes observed in the past years is fundamental to recognize which areas of the basin might be more vulnerable to climate change in order to design adaptation strategies. A general characterization of land use and livestock production of Brazilian LPB was done by using the System of Automatic Retrieving (SIDRA) of Brazilian Institute of Geography and Statistics (IBGE) platform as the major source of data. It was observed expressive increases in land areas used for temporary crops, such as soybean, sugarcane, and maize, as well as increases in poultry and swine production. These important changes in agricultural land use and livestock production are currently associated to non-climatic drivers, but this dynamic might be strongly affected by the consequences of climate change and variability, with negative socio-economic impacts for the whole region.     

Amazon forest biomass density maps: tackling the uncertainty in carbon emission estimates

As land use change (LUC), including deforestation, is a patchy process, estimating the impact of LUC on carbon emissions requires spatially accurate underlying data on biomass distribution and change. The methods currently adopted to estimate the spatial variation of above- and below-ground biomass in tropical forests, in particular the Brazilian Amazon, are usually based on remote sensing analyses coupled with field datasets, which tend to be relatively scarce and often limited in their spatial distribution. There are notable differences among the resulting biomass maps found in the literature. These differences subsequently result in relatively high uncertainties in the carbon emissions calculated from land use change, and have a larger impact when biomass maps are coded into biomass classes referring to specific ranges of biomass values. In this paper we analyze the differences among recently-published biomass maps of the Amazon region, including the official information used by the Brazilian government for its communication to the United Nation Framework on Climate Change Convention of the United Nations. The estimated average pre-deforestation biomass in the four maps, for the areas of the Amazon region that had been deforested during the 1990–2009 period, varied from 205?±?32 Mg ha^?1 during 1990–1999, to 216?±?31 Mg ha^?1 during 2000–2009. The biomass values of the deforested areas in 2011 were between 7 and 24 % higher than for the average deforested areas during 1990–1999, suggesting that although there was variation in the mean value, deforestation was tending to occur in increasingly carbon-dense areas, with consequences for carbon emissions. To summarize, our key findings were: (i) the current maps of Amazonian biomass show substantial variation in both total biomass and its spatial distribution; (ii) carbon emissions estimates from deforestation are highly dependent on the spatial distribution of biomass as determined by any single biomass map, and on the deforestation process itself; (iii) future deforestation in the Brazilian Amazon is likely to affect forests with higher biomass than those deforested in the past, resulting in smaller reductions in carbon dioxide emissions than expected purely from the recent reductions in deforestation rates; and (iv) the current official estimate of carbon emissions from Amazonian deforestation is probably overestimated, because the recent loss of higher-biomass forests has not been taken into account.     

Hybrid Model for Urban Air Pollution Forecasting: A Stochastic Spatio-Temporal Approach

Air pollution is usually driven by a complex combination of factors in which meteorology, physical obstacles, and interactions between pollutants play significant roles. Considering the characteristics of urban atmospheric pollution and its consequent impacts on human health and quality of life, forecasting models have emerged as an effective tool to identify and forecast air pollution episodes. The overall objective of the present work is to produce forecasts of pollutant concentrations with high spatio-temporal resolution and to quantify the uncertainty in those forecasts. Therefore, a new approach was developed based on a two-step methodology. Firstly, neural network models were used to generate short-term temporal forecasts based on air pollution and meteorology data. The accuracy of those forecasts was then evaluated against an independent set of historical data. Secondly, local conditional distributions of the observed values with respect to the predicted values were used to perform spatial stochastic simulations for the entire geographic area of interest. With this approach the spatio-temporal dispersion of a pollutant can be predicted, while accounting for both the temporal uncertainty in the forecast (reflecting the neural networks efficiency at each monitoring station) and the spatial uncertainty as revealed by the spatial variograms. Based on an analysis of the results, our proposed method offers a highly promising alternative for the characterization of urban air quality.     

The state of the environment in Mexico

The present work evaluates the state of the environment in Mexico based on indicators of the present status of the country’s natural resource management, social and economical conditions and anthropogenic modifications. The Mexican environment is interpreted as a spatially open system having a historical character that is essentially determined by the continual interaction between nature, society and economy. The landscape approach is followed, considering as units of territorial analysis each one of the 145 biophysical environmental units included in the national physiographic regionalization. The assessment of 16 indicators for each biophysical environmental unit was made considering their regional environmental integrity problems, the degree of disarticulation of their structure and function, and the alteration of their territorial structure, all of which determine whether or not they accomplish their environmental functions and achieve environmental stability. The classification of the state of the environment included 5 categories in 8 combinations represented in the map of the state of the environment in Mexico for the year 2008. The map shows that nearly 47.10% of the country’s surface has an environmental status ranging between unstable and critical, the problematic areas being mostly concentrated in the southeast and center of the national territory.     

Streamflow response to regional climate model output in the mountainous watershed: a case study from the Swiss Alps

Regional climate model (RCM) outputs are often used in hydrological modeling, in particular for streamflow forecasting. The heterogeneity of the meteorological variables such as precipitation, temperature, wind speed and solar radiation often limits the ability of the hydrological model performance. This paper assessed the sensitivity of RCM outputs from the PRUDENCE project and their performance in reproducing the streamflow. The soil and water assessment tool was used to simulate the streamflow of the Rhone River watershed located in the southwestern part of Switzerland, with the climate variables obtained from four RCMs. We analyzed the difference in magnitude of precipitation, maximum and minimum air temperature, and wind speed with respect to the observed values from the meteorological stations. In addition, we also focused on the impact of the grid resolution on model performance, by analyzing grids with resolutions of 50 × 50 and 25 × 25 km². The variability of the meteorological inputs from various RCMs is quite severe in the studied watershed. Among the four different RCMs, the Danish Meteorological Institute provided the best performance when simulating runoff. We found that temperature lapse rate is significantly important in the mountainous snow and glacier dominated watershed as compared to other variables like precipitation, and wind speed for hydrological performance. Therefore, emphasis should be given to minimum and maximum temperature in the bias correction studies for downscaling climatic data for impact modeling in the mountainous snow and glacier dominated complex watersheds.     

A Middle Miocene carbonate embankment on an active volcanic slope: Ilhéu de Baixo,Madeira Archipelago,Eastern Atlantic

Carbonate factories on insular oceanic islands in active volcanic settings are poorly explored. This case study illuminates marginal limestone deposits on a steep volcanic flank and their recurring interruption by deposits linked to volcaniclastic processes. Historically known as Ilhéu da Cal (Lime Island), Ilhéu de Baixo was separated from Porto Santo, in the Madeira Archipelago, during the course of the Quaternary. Here, extensive mines were tunnelled in the Miocene carbonate strata for the production of slaked lime. Approximately 10 000 m³ of calcarenite (−1 to 1ø) was removed by hand labour from the Blandy Brothers mine at the south end of the islet. Investigations of two stratigraphic sections at opposite ends of the mine reveal that the quarried material represents an incipient carbonate ramp developed from east to west and embanked against the flank of a volcanic island. A petrographic analysis of limestones from the mine shows that coralline red algae from crushed rhodoliths account for 51% of all identifiable bioclasts. This material was transported shoreward and deposited on the ramp between normal wave base and storm wave base at moderate depths. The mine's roof rocks are formed by Surtseyan deposits from a subsequent volcanic eruption. Volcaniclastic density flows also are a prevalent factor interrupting renewed carbonate deposition. These flows arrived downslope from the north and gradually steepened the debris apron westwards. Slope instability is further shown by a coral rudstone density flow that followed from growth of a coral reef dominated by Pocillopora madreporacea (Lamarck), partial reef collapse, and transport from a more easterly direction into a fore‐reef setting. The uppermost facies represents a soft bottom at moderate depths in a quiet, but shore‐proximal setting. Application of this study to a broader understanding of the relationship between carbonate and volcaniclastic deposition on oceanic islands emphasizes the susceptibility of carbonates to dilution and complete removal by density flows of various kinds, in contrast to the potential for preservation beneath less‐disruptive Surtseyan deposits. Copyright © 2013 John Wiley & Sons, Ltd.     

Multifrequency algorithms for precise point positioning: MAP3

We present the new MAP3 algorithms to perform static precise point positioning (PPP) from multifrequency and multisystem GNSS observations. MAP3 represents a two-step strategy in which the least squares theory is applied twice to estimate smoothed pseudo-distances, initial phase ambiguities, and slant ionospheric delay first, and the absolute receiver position and its clock offset in a second adjustment. Unlike the classic PPP technique, in our new approach, the ionospheric-free linear combination is not used. The combination of signals from different satellite systems is accomplished by taking into account the receiver inter-system bias. MAP3 has been implemented in MATLAB and integrated within a complete PPP software developed on site and named PCube. We test the MAP3 performance numerically and contrast it with other external PPP programs. In general, MAP3 positioning accuracy with low-noise GPS dual-frequency observations is about 2.5 cm in 2-h observation periods, 1 cm in 10 h, and 7 mm after 1 day. This means an improvement in the accuracy in short observation periods of at least 7 mm with respect to the other PPP programs. The MAP3 convergence time is also analyzed and some results obtained from real triple-frequency GPS and GIOVE observations are presented.