Analysis of Meteorological data using Python

Objective:

     The goal is to analyse the Meteorological dataset,study the underlying features and relationships between the variables to analyse the weather changes using python.

Dataset:

The Dataset used is an extract from Meteorological Dataset Kaggle.

Overview of the Dataset:

Features:

• Formatted Date(in datetime format)
• Summary(weather type)
• Precip type(4 values)
• Temperature(C) (in celsius)
• Apparent Temperature(C) (in celsius)
• Humidity (in grams per kilogram)
• Wind Speed(km/h) (max wind speed achieved)
• Wind Bearing (degrees) (direction of the wind)
• Visibility(km) (amount of the visibility)
• Loud Cover(null values)
• Pressure(millibars) (pressure of the wind)
• Daily Summary(overview of  the weather)

Data information:

• Columns-count : 12
• Row-count : 96,453
• Quantitative features[8] : temperature , apparent temperature , humidity , wind speed , wind bearing , visibility , loud cover , pressure.
• Discrete (Categorical) features[4] : formatted date , summary , precip type , daily summary.

-After Resampling the data from hourly to monthly we can visualize the data using matplotlib and seaborn library

Visualizing Data:

        Quantitative Variables:

This is the Histogram representation of the Quantitative data.From the above graph it is clear that Loud Cover column has all null values , so we have to drop this 

column as it is unnecessary.

         Categorical Variables:

                                   

This graph shows the countplot of the categorical representation of the Summary of the weather achieved.Over 10 years most of the days have achieved cloudy weather.

Analysis of Humidity vs Apparent Temperature in the month of April:

In the month of april over all the years from 2006 to 2016 we can analyse that there is a normal measure of Humidity and Apparent Temperature has increased more in the year of 2009 , from 2012 to 2014 there is slight increase in app. temperatue and decreased in 2015.

Variation or change of the Humidity in all months from 2006 to 2016 is depicted below:

The above graph shows that there is more humidity. It is clear that the humidity has reached it's peak on 2011,2016, and drastic fall on 2007,2012.It is obvious that November,December and January are the most humid months and one can expect rainy season during this period.

Changes in Apparent temperature of different months from 2006 to 2016 can be visualized as:                                      

It shows that there is more temperature on july month for all the years and december is the perfect month for the winter.But april and october are the months where we will feel the weather with moderate temperature.

Variation in Average Humidity in the month of april over the years:                                                                           

                   We can see that Humidity changing drastically over the years and 2006 is the year where we experience the humidity of 0.730 and fallen to 0.535 in 2007.

Variation in Average Apparent temperature in the month of April for the years:       

We can observe that apparent temperature of the month april has increased rapidly in the year of 2009 and fallen to the range of 10.6 to 10.8 in 2015.

Conclusion:

                   From the above analysis,using variation in various factors such as Humidity,Apparent Temperature ,Wind speed and pressure-

• We can observe the changes in a weather of particular location , Seasonal changes using Python analysis.
• This analysis helps Meteorologists to observe,understand,predict and forecaste the weather phenomena.
• We can observe that Humidity is high in winter months and Apparent Temperature in june and july.Through this,we can predict the upcoming weather reports using Python data analysis and visualization.

 

For analysis:

 https://www.kaggle.com/saishankargyenagi/performing-analysis-of-meteorological-data

        

Recommended:

 https://datascience.suvenconsultants.com