Search

In recent years there has been a renewed interest in utilitarian cycling due to its recognized potential in the reduction of energy consumption and pollution in the cities. Bike-sharing systems generate a large amount of data which can be used to improve the systems themselves, or to improve the body of knowledge on urban mobility. The open data automatically generated by the Brussel's bike-sharing system (Villo) is explored through spatial regression models of the number of bicycle trips at stations. The main goal of the modelling process is to understand if socio-economic, infrastructure and land use factors influence mobility patterns in peak periods and weekdays. The first step of the modelling process consists in setting up exploratory Ordinary Least Squares (OLS) models in order to identify potential explanatory variables. Finally, Geographically Weighted Poisson Regression (GWPR) models and semi-parametric versions of GWPR models are parametrised using the previously identified variables. The results show that the relationships between the dependent and independent variables are complex and spatially varying. Furthermore, the results show hidden patterns that enable further local investigation on these relationships. The weaknesses and strengths of our approach are discussed, particularly its implementation in other geographic contexts and its potential of generalisation for all bicycle trips. [ABSTRACT FROM AUTHOR]

Air temperature influences a variety of environmental processes, having a significant impact on the conditions of living of humans and other life forms. The Brazilian Northeast is a region that comprises a diversity of ecosystems, but it is most known as a semi-arid area characterized by severe environmental conditions. This work proposes to model air temperature in Brazilian Northeast to evaluate changing patterns between 2000 and 2017 using two interpolation techniques and comparing them. Monthly average temperature data from meteorological stations were gathered and used to compute the average annual temperature. Then, the timeframe was divided into 2 periods: 2000 to 2008 (1) and 2009 to 2017 (2), and the average temperature of each period was computed based on the annual average. Descriptive statistics analysis and exploratory spatial data analysis were performed, providing insights on the temperature patterns and distribution. In addition, interpolated surfaces were generated using the Inverse Distance Weighting and Ordinary Kriging methods for each period, and results were compared using error statistics derived with cross-validation. The results revealed that for both periods the highest temperatures are exhibited in the northern and central regions, whereas the lowest values occur in the south and east. In terms of change, an overall increase in the average temperature was noticed from period 1 to 2, although in some areas the increase was greater than in others. There was an increase of 0,37ºC in the mean, 1,37ºC in the maximum and 0,18ºC in the minimum temperature over the study region. Furthermore, Ordinary Kriging produced better results in terms of the bias of the predictions. The interpolated surfaces allow to visually notice the change in the average temperature between the periods. This study contributes to a better understanding of the temperature variability in the Brazilian Northeast in the 21st century. [ABSTRACT FROM AUTHOR]

Climate data homogenisation is of considerable relevance for a broad range of studies and applications. Discontinuities in the time series may bias the conclusions of those studies, so they should be detected and corrected. This document presents a complete workflow for the homogenisation of climate data, using a process named gsimcli, which is based on a geostatistical simulation method. The devised workflow is illustrated with a benchmark data set. It is expected that this contribution will help technicians, researchers, and other professionals to detect and to correct irregularities in climate data. [ABSTRACT FROM AUTHOR]

Remote sensing data and imagery is used in countless applications, but their analysis and interpretation highly depends on the spatial and spectral resolution. Fusion has been defined as the process of combining relevant image(s) information in order to synthesize it into more informative and more suitable image(s) for visual perception or computer processing, particularly for interpretation or classification. Building upon these objectives, this study explores the application of ordinary cokriging and collocated cokriging to improve the spatial resolution of the multispectral bands from a Landsat ETM+ image based on its corresponding panchromatic image. The study region, located in the central part of Colombia, has different land uses and covers such as water, forest, crop and urban areas. The geostatistical methods allowed for the fusion of the Landsat's original bands into 'fused bands', which improved the visualization and characterization of the study area. The collocated cokriging results were slightly less satisfactory than those of ordinary cokriging, possibly because of the land cover heterogeneity of the region. The results indicate that the fused bands obtained through this generally available technique can provide valuable information in digital processing of Landsat images, such as land use and land cover classification. [ABSTRACT FROM AUTHOR]

In the Iberian Peninsula, precipitation mainly falls during autumn and winter and is marked by large space and time variability. Therefore, the Peninsula has large areas prone to drought. This study aims to analyse spatiotemporal drought patterns in the Iberian Peninsula based on the Vegetation Condition Temperature Index (VTCI) and to investigate the response of land cover types towards droughts during spring and summer seasons from 2001 to 2005 and in 2007 and 2010 (highly dry and wet years). VTCI is a remote sensing drought index and uses information on vegetation activity (Normalized Differenced Vegetation Index - NDVI) and Land Surface Temperature (LST). VTCI overcomes problems that may arise when NDVI is solely used for drought monitoring. The inter-annual VTCI course shows a typical Mediterranean pattern that is particularly influenced by the relief. The land cover types most prone to drought are agricultural surfaces, and the most resilient ones are forests. [ABSTRACT FROM AUTHOR]