Cloud based spatial visualization with statistical approach for road accidents Kumari Pritee 1 • R. D. Garg 2 Received: 18 July 2017 / Revised: 13 November 2017 / Accepted: 15 November 2017 Ó Korean Spatial Information Society 2017 Abstract Road accidents have emerged as a major public health problem globally. The scope of road accident data usage is limited to serve statistical purposes. Therefore it is necessary to identify high-frequency accident locations by performing risk evaluation using Geographic Information System (GIS) and statistical analysis to enhance road safety. GIS is a powerful spatial analytics system for accident pattern analysis. Our aim is to provide a cloud based GIS application for Roorkee City that visualizes the analysis of hotspots in raster as well as vector format using kernel density estimation, buffer analysis and nearest neighborhood analysis for accident and hospital locations over the cloud to achieve shortest path from accident locations to nearest hospitals and police stations. The accident locations have been classified into five zones i.e. very high accident frequency zone, accident frequency zone, moderate accident frequency zone, less accident frequency zone and very less accident frequency zone using heat map analysis with the help of kernel density estimation. Quantum-GIS have been applied for heatmap analysis, buffer analysis and nearest neighborhood analysis over the cloud. Periodically the database will get updates and reduce accident severity. Any user can get any infor- mation about road accident attributes with coordinates, date and time over the cloud and will be of benefit for decision-making processes. Keywords SPSS Á Internet-GIS Á Spatial operations Á Quantum-GIS Á Excel Á Kernel methods 1 Introduction Road accidents are increasing day by day contributing major deaths and severe injury. According to the official website of NHAI (National Highways Authority of India), almost 5 lakh accidents occurred in India in 2015 in which approx. 146,000 people dead and leaving thrice the number injured. India has a 3.3 million km road network including national highways (NH), state highways (SH), major dis- trict roads (MDR) and other district roads (ODR), which makes the country the second largest road network in the world. 2% of the total road length occupied by NH carries 40% of passenger traffic and 85% of goods traffic. 20% of the accidents on NH have occurred due to goods traffic [1]. 430,654 people lost their lives due to road traffic crashes in 2010 in India [2]. The situation got worse in recent years. From year 2009 to 2010 traffic fatalities have increased by about 5.5% per year because of an increase in the number of vehicles on the road, and the absence of a coordinated official policy to control the problem i.e. 443 and 367 deaths per day and 1301 and 1290 injuries per day due to traffic accidents and road accidents respectively, and 73 deaths per day by trucks/lorries and 77 deaths by two- wheelers [2]. By collecting attribute data (Spot speed of the vehicles at different locations during daytime and nighttime, traffic volume, road width, surface conditions, weather conditions etc.) from field survey and police stations, different Electronic supplementary material The online version of this article (https://doi.org/10.1007/s41324-017-0148-9) contains supple- mentary material, which is available to authorized users. & Kumari Pritee priteegeo.kumari23@gmail.com 1 Centre for Transportation Systems (CTRANS), IIT Roorkee, Roorkee, India 2 Geomatics Engineering, IIT Roorkee, Roorkee, India 123 Spat. Inf. Res. DOI 10.1007/s41324-017-0148-9