AffectRoute – considering people’s affective responses to environments for enhancing route-planning services

Humans perceive and evaluate environments affectively. Some places are experienced as unsafe, while some others as attractive and interesting. These affective responses to environments influence people’s daily behavior and decision-making in space, e.g., choosing which route to take, or which place to visit. In this article, we report on a methodology of using people’s affective responses to environments for enhancing computer-based route planning. More specifically, we explore a crowdsourcing approach to model and collect people’s affective responses to environments; an Affect-Space-Model and a mobile application are developed to facilitate this crowdsourcing approach; a routing algorithm (named AffectRoute) is then proposed to aggregate and integrate the collected data for automatic route planning. Evaluation with human participants shows that the routes generated by considering people’s affective responses to environments are significantly preferred over the conventional shortest ones, which are employed in car navigation systems and many online route planners. In conclusion, considering people’s affective responses to environments contributes to the improvement of automatic route planning. The proposed method can be integrated into existing route-planning services (e.g., location-based services) to provide users with more satisfying routing results.     

Thermal bioclimate in idealized urban street canyons in Campinas, Brazil

Among several urban design parameters, the height-to-width ratio (H/W) and orientation are important parameters strongly affecting thermal conditions in cities. This paper quantifies changes in thermal comfort due to typical urban canyon configurations in Campinas, Brazil, and presents urban guidelines concerning H/W ratios and green spaces to adapt urban climate change. The study focuses on thermal comfort issues of humans in urban areas and performs evaluation in terms of physiologically equivalent temperature (PET), based on long-term data. Meteorological data of air temperature, relative humidity, wind speed and solar radiation over a 7-year period (2003–2010) were used. A 3D street canyon model was designed with RayMan Pro software to simulate the influence of urban configuration on urban thermal climate. The following configurations and setups were used. The model canyon was 500 m in length, with widths 9, 21, and 44 m. Its height varied in steps of 2.5 m, from 5 to 40 m. The canyon could be rotated in steps of 15°. The results show that urban design parameters such as width, height, and orientation modify thermal conditions within street canyons. A northeast–southwest orientation can reduce PET during daytime more than other scenarios. Forestry management and green areas are recommended to promote shade on pedestrian areas and on façades, and to improve bioclimate thermal stress, in particular for H/W ratio less than 0.5. The method and results can be applied by architects and urban planners interested in developing responsive guidelines for urban climate issues.     

Knowledge and Reasoning in Spatial Analysis

Reasoning is an essential part of any analysis process. Especially in visual analytics, the quality of the results depends heavily on the knowledge and reasoning skills of the analyst. In this study, we consider how to make the results transparent by visualizing the reasoning and the knowledge, so that persons from outside can trace and verify them. The focus of this study is in spatial analysis and a case study was carried out on a process of off‐road mobility analysis. In the case study, linked views of a map and a PCP were identified as reasoning artifacts. The knowledge used by the analyst was formed by these artifacts and the tangible pieces of information identified in them, along with the mental models of the analyst′s mind. To make the results transparent, the tangible pieces of information were marked with sketches and the mental models were presented in causal graphs because it was found that causality was central to the reasoning process in the case study. The causal graph allows the reasoning of the analyst to be studied, as well as traced back to its origin.     

The role of the simulation setup in a long-term high-resolution climate change projection for the southern African region

High-resolution regional climate change simulations have proven to offer an added value compared to available global climate model simulations. However, over many regions of the globe, long-term high-resolution climate change projections are rather sparse. We present a transient high-resolution climate change projection with the regional climate model with the regional climate model REMO over the southern African region, following the SRES A1B emission scenario. The simulation was conducted at 18?km grid spacing for the period from 1960 to 2100, making it to the longest available climate change projection at such a high resolution for the region. In the first part of the study, we focus on the impact of the model setup on the simulated rainfall over the southern African region. In the standard setup, we used the output of the global climate model ECHAM5/MPIOM directly to force REMO. This setup led to a very strong wet bias over the region. Changing it to the double-nesting setup significantly reduced this bias, but a substantial wet bias still persists. The remaining bias could partly be attributed to a warm bias in the SST forcing over the southern Atlantic Ocean. Thus, we applied an SST correction based on the anomaly approach to the data, which led to a further improvement of the rainfall simulation. As the SST bias in the southern Atlantic is a common feature of all global climate models assessed by the IPCC, we recommend the chosen model setup, including the SST correction, as general procedure for dynamical downscaling studies over the southern African region. In the second part, we present the projected spatial and temporal changes of temperature and precipitation, including several rainfall characteristics, over the southern African region. Herby we compare the projections of the high-resolution REMO simulation to those of the forcing regional and global models. We generally find that for temperature the magnitude of the projected changes of the regional model only slightly differs from the GCM projection; however, the spatial patterns are much better resolved in the RCM projections. For precipitation, REMO shows a more intense drying toward the end of the twenty-first century than it is simulated by the global model. This can have a major influence when investigating the impacts of future climate change on a regional or even local scale. In combination with the improved spatial patterns, the application of high-resolution climate change information could therefore improve the results of such applications.     

How dynamical models can learn from the data��an example with a simplified ENSO model

Most dynamical models of the natural system contain a number of empirical parameters which reflect our limited understanding of the simulated system or describe unresolved subgrid-scale processes. While the parameterizations basically introduce some uncertainty to the model results, they also hold the prospect of tuning the model. In general, a deterministic tuning is related to an inversion of the model which is often impossible or requires considerable computing effort for most climate models. Another way to adjust the model parameters to a specific observed process is stochastic fitting where a set of parameters and model output are taken as random variables. Here, we present a dynamical?Cstatistical approach with a simplified model of the El Ni?o?CSouthern Oscillation (ENSO) cycle whose parameters are adjusted to simulated and observed data by means of Bayesian statistics. As ENSO model, we employ the Schop?CSuarez delay oscillator model. Monte Carlo experiments highlight the large sensitivity of the model results to varied model parameters and initial values. The statistical adjustment is done by Bayesian model averaging of the Monte Carlo experiments. Applying the method to simulated data, the posterior ensemble mean is much closer to the reference data than the prior ensemble mean. The learning effect of the model is evident in the leading empirical orthogonal functions and statistically significant in the mean state. When the method is applied to the observed ENSO time series, the ENSO model in its classical setup is not able to account for the temporally varying periodicity of the observed ENSO phenomenon. An improved setup with continuous adjustment periods and extended parameter range is developed in order to allow the model to learn from the data gradually. The improved setup leads to promising results during the twentieth century and even a weak forecast skill over 6?months. Thus, the described method offers a promising tool for data assimilation in dynamical weather and climate models. However, the simplified ENSO model is barely appropriate for operational ENSO forecasts owing to its limited physical complexity.     

Probabilistic assessment of regional climate change in Southwest Germany by ensemble dressing

Since single-integration climate models only provide one possible realization of climate variability, ensembles are a promising way to estimate the uncertainty in climate modeling. A statistical model is presented that extracts information from an ensemble of regional climate simulations to estimate probability distributions of future temperature change in Southwest Germany in the following two decades. The method used here is related to kernel dressing which has been extended to a multivariate approach in order to estimate the temporal autocovariance in the ensemble system. It has been applied to annual and seasonal mean temperatures given by ensembles of the coupled general circulation model ECHAM5/MPI-OM as well as the regional climate simulations using the COSMO-CLM model. The results are interpreted in terms of the bivariate probability density of mean and trend within the period 2011–2030 with respect to 1961–1990. Throughout the study region one can observe an average increase in annual mean temperature of approximately +0.6K and a corresponding trend of +0.15K/20a. While the increase in 20-year mean temperature is virtually certain, the 20-year trend still shows a 20% chance for negative values. This indicates that the natural variability of the climate system, as far as it is reflected by the ensemble system, can produce negative trends even in the presence of longer-term warming. Winter temperatures are clearly more affected and for both quantities we observe a north-to-south pattern where the increase in the very southern part is less intense.     

Ocean circulation on the North Australian Shelf

The ocean circulation on Australia's Northern Shelf is dominated by the Monsoon and influenced by large-scale interannual variability. These driving forces exert an ocean circulation that influences the deep Timor Sea Passage of the Indonesian Throughflow, the circulation on the Timor and Arafura Shelves and, further downstream, the Leeuwin Current. Seasonal maxima of northeastward (southwestward) volume transports on the shelf are almost symmetric and exceed 10⁶ m³/s in February (June). The associated seasonal cycle of vertical upwelling from June to August south of 8.5°S and between 124°E and 137.5°E exceeds 1.5×10⁶ m³/s across 40 m depth. During El Niño events, combined anomalies from the seasonal means of high regional wind stresses and low inter-ocean pressure gradients double the northeastward volume transport on the North Australian Shelf to 1.5×10⁶ m³/s which accounts for 20% of the total depth-integrated transport across 124°E and reduce the total transport of the Indonesian Throughflow. Variability of heat content on the shelf is largely determined by Pacific and Indian Ocean equatorial wind stress anomalies with some contribution from local wind stress forcing.     

Characterization of permeability and porosity from nanosensor observations

In this paper, we investigate the information content in “nanosensors” with limited functionality that might be injected into a reservoir or an aquifer to provide information on the spatial distribution of properties. The two types of sensors that we consider are sensors that can potentially measure pressure at various times during transport, and sensors can be located in space by perturbations in electrical, magnetic, or acoustic properties. The intent of the study is to determine the resolution of estimates of properties that can be obtained from various combinations of sensors, various frequencies of observations, and various specifications on sensor precision.Our goal is to investigate the resolution of model estimates for various types of measurements. For this, we compute linearized estimates of the sensitivity of the observations to the porosity and permeability assuming gaussian errors in the pressure and location observations. Because the flow is one-dimensional and incompressible, observations of location are sensitive to the porosity between the injection location and the sensor location, while the location of particles is sensitive to the effective permeability over the entire interval from injector to producer. When only the pressure is measured but the location of the sensor is unknown, as might be the situation for a threshold sensor, the pressure is sensitive to both permeability and porosity only in the region between the injector and sensor.In addition to the linearized sensitivity and resolution analyses, Markov chain Monte Carlo sampling is used to estimate the posterior pdf for model variables for realistic (non-Gaussian) likelihood models. For a Markov chain of length one million samples approximately 200-500 independent samples are generated for uncertainty and resolution assessment. Results from the MCMC analysis are not in conflict with the linearized analysis.