PUI2016 - Extra Credit Project<Avikal Somvanshi, as10724, as10724>Abstract: Ground-level ozone is a deadly pollutant and even short term exposure is know to cause health impact. But unlike other aerosols, ozone is not emitted directly into the air from any one source, but is created by photo-chemical reactions among oxides of nitrogen (NOx) in the presence of sunlight. Therefore, to control ozone pollution it is critical to understand the process of its formation in real world and what role weather and various pollutants play in this process. This study analysed hourly pollution and weather data from the Delhi, India from April 2015 to November 2016 to answer these questions. The results show that ozone concentration in Delhi is correlated with solar radiation, temperature, relative humidity and NOx concentration. No correlation was found with wind speed and PM2.5 concentration. Study also found that instances of high ozone pollution are distributed throughout the year (only exception being rainy days during the monsoon season) in Delhi and it is not only a summer season phenomena as generally believed.
PUI2016 Extra Credit Project ProposalAnalysis of Ground-level Ozone formation in Delhi, India <Avikal Somvanshi, as10724, as10724>Problem Description: Ground-level ozone is a deadly secondary pollutant, that it is not emitted directly into the air from any one source, but is created by chemical reactions between oxides of nitrogen (NOx) and volatile organic compounds (VOC) in the presence of sunlight. Analysis will estimate impact of ambient temperature and solar radiation on ground-level ozone formation while comparing relative change in concentration of NOx and Ozone in the atmosphere.Research Question: Is ground-level ozone pollution is linked to concentration of oxides of nitrogen (NOx), ambient temperature and solar radiation? Data: Hourly pollution (Particulate Matter 2.5, Ozone and Oxides of Nitrogen) and weather (temperature, wind speed, relative humidity and solar radiation) data from two Pollution Monitoring stations of the Delhi government from April 2015 to Nov 19, 2016. One station is in South Delhi, while other in West Delhi, about 10 miles apart. Mined from http://www.cpcb.gov.in/CAAQM/frmUserAvgReportCriteria.aspx Hourly data from Delhi, one of the most polluted urban air in the world, is best suited for this analysis as it gives good data resolution both in timeline and observed concentration. Analysis: The project will identify events of formation of ground-level ozone through outlier detection via thresholding and will check for patterns using the Fourier analysis. Impact of NOx concentration, ambient temperature and solar radiation will be assessed via linear regression analysis.References: Delhi Summer Ozone Data Sheet , 2014, Centre for Science and Environment, http://www.cseindia.org/userfiles/ozone-fact-sheet-20140609.pdfZhao H, Wang S, Wang W, Liu R, Zhou B (2015) Investigation of Ground-Level Ozone and High-Pollution Episodes in a Megacity of Eastern China. PLoS ONE 10(6): e0131878. doi:10.1371/journal.pone.0131878, http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0131878Bordignon, S., Gaetan, C. & Lisi, F., Nonlinear models for ground-level ozone forecasting: Statistical Methods & Applications (2002) 11: 227. doi:10.1007/BF02511489, http://link.springer.com/article/10.1007/BF02511489Ambient (outdoor) air quality and health factsheet, WHO, http://www.who.int/mediacentre/factsheets/fs313/en/2015 National Ambient Air Quality Standards (NAAQS) for Ozone, US EPA, https://www.epa.gov/ozone-pollution/2015-national-ambient-air-quality-standards-naaqs-ozoneOzone pollution, US EPA, https://www.epa.gov/ozone-pollutionDeliverable: The project will establish a statistical and graphical tool to predict ground-level ozone formation based on NOx and weather data. This will help city agencies to issue preventive health advisories to the people and undertake corrective action.
Abstract: The study analyses the CitiBike usage data to compare trip duration among men and women. This was done using Citibike open data for January 2015. The analysis found out that women on average rent out citibikes for longer duration than men, especially women in higher age bracket compared to men in same age bracket. Data: The study uses the usage data from January 2015. The data was sorted by age, by trip duration and by gender. Entries marked as anything other than male or female in gender section was dropped from the study.Hypothesis: Women bike for longer duration than men. NULL HYPOTHESIS: The average trip duration of women biking is the same or lower than the average trip duration of a woman biking. Ho : Fduration.mean() <= Mduration.mean() Ha : Fduration.mean() > Mduration.mean() Alpha=0.05Analysis: Since the study compares two independent set of data set, T-test was used to check the validity of the null hypothesis. Result: The t-test return a pvalue that was less than alpha assumed for the study therefore the Null Hypothesis does not hold and is rejected. Which mean that women on average take out citibikes for long duration than men. Further, difference between mean trip duration of men and women is 1.16364964914 minutes. This further establishes that women on average ride bikes for longer duration than men. Github file: https://github.com/as10724/PUI2016_as10724/blob/master/HW3_as10724/Assignment%202.ipynbFigure1: