Search

Machine learning (ML) interpretability has become increasingly crucial for identifying accurate and relevant structural relationships between spatial events and factors that explain them. Methodologically aspatial ML algorithms with an apparent high predictive power ignore non-stationary domain relationships in spatio-temporal data (e.g. dependence, heterogeneity), leading to incorrect interpretations and poor management decisions. This study addresses this critical methodological issue of 'interpretability' in ML-based modeling of structural relationships using the example of heterogeneous drivers of wildfires across the United States. Specifically, we present and evaluate a spatio-temporally interpretable random forest (iST-RF) that uses spatio-temporal sampling-based training and weighted prediction. Although the ultimate scientific objective is to derive interpretation in space-time, experiments show that iST-RF can improve predictive accuracy (76%) compared to the aspatial RF approach (70%) while enhancing interpretations of the trained model's spatio-temporal relevance for its ensemble prediction. This novel approach can help balance prediction and interpretation with fidelity in a spatial data science life cycle. However, challenges exist for predictive modeling when the dataset is very small because in such cases locally optimized sub-model's prediction performance can be suboptimal. With that caveat, our proposed approach is an ideal choice for identifying drivers of spatio-temporal events at country- or regional-scale studies. [ABSTRACT FROM AUTHOR]

Wrong-way crashes on limited access highways continue to be a serious problem for state highway agencies and communities. These crashes are more likely to result in fatalities or serious injuries than other traffic incidents, due to vehicles travelling at high speeds, often involved on a roadway with limited vertical space and time to avoid a crash. This research study focused on wrong-way crashes that occurred on highways in the State of Kansas. Although these crashes represent a very small portion of crashes, wrong-way crashes were found to have a higher rate of fatalities and injuries as compared to other crash types. Using ten years of crash data, it was found that a typical wrong-way crash occurred under no adverse weather conditions, at a non-intersection location, in daylight or with streetlights present, involved alcohol or drugs, and resulted in a fatality or serious injury. An ordinary logistic model was developed to identify significant characteristics of wrong-way crashes. The model indicated that drivers under the influence of alcohol or drugs was found to be a significant in both fatal and injury wrong-way crashes. It was also found that certain lighting conditions were also significant, along with use of safety equipment and drivers over the age of 55 years old. Additional research is needed to further investigate wrong-way crashes and their causalities in Kansas on other roadway facilities including rural at-grade divided intersections. [ABSTRACT FROM AUTHOR]

We propose a novel approach to network-based event likelihood modeling that estimates the probabilities of event occurrences on a network and identifies the influences of site and situation characteristics. Our premise is that the occurrences of events that involve human activities are subject to site and situational characteristics, and an understanding of event occurrences serves the basis for preparation or mitigation. Using data from Dallas, Texas, USA, we take the proposed approach to estimate the likelihood of traffic accidents based on binary (event or nonevent) space-time atoms of 100-m road segments and 1-h intervals. We choose 12 variables representing time, site characteristics, and situational conditions based on literature reviews to develop logistic regression and random forest models. The traffic accident data on even days were used for model construction and data on odd days for model testing. Both models result in comparable accuracy at 84.11% (logistic regression) and 85.42% (random forest) with significant differences in the spatial patterns of how site and situation correlate to traffic accidents. The difference signals the dynamic influence of site and situation characteristics on the event likelihood over time. The proposed approach shall be applicable to other point events on a network. [ABSTRACT FROM AUTHOR]

Study abroad offers a specific configuration of encountering cultural Others: in a demarcated “abroad” space in a compartmentalized time period. Informed by Bakhtin's notion of chronotopes, this article investigates narratives of American college students who studied in Paris, France and Bilbao, Spain in June–July 2011 based on an ethnographic fieldwork of their stay. I identify two chronotopes held in tension in their narratives that reflect tenets of the discourse of immersion: of homogeneous space where every minute students spend in the host society is “local” time full of learning, which risks portraying the host society as frozen in the time of the students’ stay; and of heterogeneous space where local space–time and outsider (e.g. American students and tourists) space–time co-exist hierarchically, where students strive to show their engagement with the former. This article calls for encouraging students to examine the effects of these chronotopes on their experiences and for viewing study abroad not as an encounter of two cultures but as diverse students joining in the ongoing production of heterogeneous host society space with a compartmentalized yet expanding notion of time. [ABSTRACT FROM PUBLISHER]

The idea that statistical relationships can change over time and space has been around for many centuries. Social scientists began investigating how statistical relationships shift over time, with techniques such as time series analysis, long ago and have only recently begun to explore shifts over geographical space. A map of the USA. mainland that makes use of recent population data to regress a geographically weighted regression is presented to show the reader an instance when a statistical relationship varies as a function of geographical space. [ABSTRACT FROM AUTHOR]

Time-space compression shows students that geographies are plastic, mutable and forever changing. This paper justifies the need to teach this topic, which is rarely found in undergraduate course syllabi. It addresses the impacts of transportation and communications technologies to explicate its dynamics. In summarizing various conceptual perspectives, the paper illustrates the diverse ways in which it has been theorized. The links between time-space compression and globalization are explored. The paper points to four common pitfalls that plague attempts to teach this issue. It offers five classroom examples through which time-space compression can be taught. [ABSTRACT FROM AUTHOR]

Central to temporal Geographical Information systems (GIS) research is representation and analysis of temporality in spatial databases. Temporality can be characterized by change, motion, and periodicity, but to date, most temporal GIS research advances contribute to representing and analyzing spatially discrete constructs of amandant vectors, geospatial lifelines, space-time trajectories, events and network flows, none of which accounts for geographic phenomena that are continuously dispersing or propagating in space and time. This article introduces a temporal GIS framework that adopts concepts of fluid kinematics and uses velocity as the basic unit of spatiotemporal representation to identify kinematic flows and features indicative of geographic processes, such as divergence and convergence. Climate projections of IPCC A2 scenario are used to test the proposed kinematic representation by comparing data outputs from two popular General Circulation Models (GCMs): the Center National Weather Research global ocean-atmosphere-coupled system (CNRM-CM3) in France and the National Center for Atmospheric Research Community Climate System Model (NCAR-CCSM3) in the United States. The case study shows that fluid kinematics approach offers an alternative means to time-stamped, change-based or event-based representations of geographic phenomena and provides a new foundation for GIS analysis of space-time fields. Specifically, the proposed GIS framework enables direct manipulation of velocity to reveal flows. The case study identifies distinct regions of rapid isotherm movement. By comparison, both GCMs identify similar regions of convergence and divergence except for the following: a high-latitude region in southern Greenland, a band of regions at -15° latitude extending from the middle of South America to east of Madagascar, a region over the Arabian Sea, a region over Indonesia, and several regions in the tropical oceans. [ABSTRACT FROM AUTHOR]

Based upon early ethnographic accounts, the ice-glider was a recreational artifact, with limited distribution in time and space, first appearing in the early historic period and confined to the Missouri River drainage. Recent research has shown a more broad distribution across both time and space, extending its use into the precontact period and extending its distribution northward into the Northern Plains and the adjacent Parklands of Canada. Its earliest dated occurrence is in the parklands of southern Manitoba and the southern plains of Saskatchewan circa 1440 A.D. Ice-gliders provide a glimpse of non-subsistence activities of First Nations people prior to the advent of Europeans in the region. They also provide a useful indicator of site seasonality. Elaborate decoration on some specimens suggests to the authors that the ice-glider may have had non-secular applications as well. [ABSTRACT FROM AUTHOR]

In this essay we examine the metaphorical rendition of the terrorist attacks on New York and Washington, which took place on 11 September 2001, as attacks on "civilization." Our principal aim is to understand how it has proven to be as affective as it seems to have been thus far. We do so by turning to film, a medium that is quintessentially metaphorical and whose own affective power rests on the reconfiguration of time and space. More specifically, we do so by turning to David Fincher's Fight Club (1999). In contrast to the textual approach offered by Slavoj Žižek (2002), we explore how the formal play of different temporal and spatial scales, which would articulate a trenchant critique of alienation, in fact reproduces an understanding of the subject that is deeply complicit with capital. The play of time and space, on which the affective power of the film rests, articulates a rejection of the fundamentally schizophrenic nature of the subject of capital. Our claim is that this exploration of time, space, and political critique illuminates how the elisions of the historical and geo-political that characterize the metaphorical rendition of the attacks on New York and Washington might easily be overlooked. This essay also advances our more general understanding of both the articulation of time and space within the dominant form of metaphorical representation and how this enacts and consolidates a particular politics. [ABSTRACT FROM AUTHOR]