Session 13 - Geographical information systems (GIS)

Abstract: The treatment of health emergencies involves several quantifiable processes. In this context, this article proposes a quantitative methodological tool that seeks to determine the best scenarios in order to obtain an efficient reduction in response times. Within the services of pre-hospital care, transportation plays a critical role in the comprehensive management of medical emergencies, so, determining the ideal location of the place from where the ambulances meet the demand, influenced the time of arrival of the resource. Getting to determine the best location of ambulances for health management can be a complex task, given the number of variables involved in this process, and the financial costs involved in conducting empirical experimentation in the field. The proposed model sought to provide decision makers with quantitative probabilistic tools that allowed obtaining experimental results while minimizing implementation costs, through the operationalization of demand and availability variables and location of ambulances based on their distribution of probabilities, which , through stochastic processes, they determined the impact on the time of arrival of the resource, this reduction of the arrival time of the ambulance, was contrasted with the initial data arriving to determine that these variations had statistical significance through parametric tests. 

Authors: Edison Roberto Valencia Nuñez, Hamilton Vinicio Montenegro Lopez and Lorenzo Jeovany Cevallos Torres

Abstract: The role of eGovernment and geographic information technology is crucial in the resolution of environmental problems such as Amazon deforestation. Yet, the assessment of deforestation based on Geographic Information Systems is highly dependent on the availability of accurate geospatial information. In this study, a Random Forest modeling approach is developed using open data and free software to identify spatially explicit deforestation drivers and areas vulnerable to forest transitions in the department of San Martín, Peru. The results of this study show that it was possible to explain 88% of the variance in deforestation based on open geospatial data, and to generate a deforestation vulnerability map with an overall classification accuracy of 89%. The modeling approach proposed in this study could aid local governments in remote tropical areas, coping with limited data availability and financial resources, to examine deforestation trends and target green government policies.

Authors: Vincent Bax