/Climate-Change-Through-Years

University Project about Climate Change from 1955 to 2020 using Python and Machine Learning techniques

Primary LanguageJupyter Notebook

Climate Change through Years

Q1: Obtain the Data

Data downloaded from the National Oceanic and Atmospheric Administration's National Centers for Environmental Information (https://www.ncdc.noaa.gov/cdo-web/) and in particular https://www.ncdc.noaa.gov/cdo-web/search.

Q2: Deviation of Summer Temperatures

Graph showing the mean summer temperature deviation from a baseline 1974-1999.

Q3: Evolution of Daily Temperatures

We will get the average temperature for each year for the full period from 1955 to 2020 and then create a plot showing the daily temperature for each year. The line corresponding to each year will be smoothed by using a 30 days rolling average. The lines are colored from light orange to dark orange, progressing through the years in ascending order.

Q4: Extreme Temperature Events

On that plot we will overlay a line showing the average daily temperature for the baseline period of 1974-1999 (that is the black line). The line will also be smoothed usng a 30 days rolling average. Another measure used by climatologists is the number of extreme events. Extreme events are defined as those beyond 5% or 10% from the expected value. We will deal with extreme heat events going 10% above the baseline of 1974-1999.The vertical axis is the percentage of extreme heat events calculated over the number of observations for each year. The gray line in the middle is the average percentage of extreme tempearture events of the baseline. The colour blue is used for those years where the percentage is below the baseline; otherwise the colour is orange.

Q5: Precipitation

Continuing the thread on extreme events, another consideration is rainfall. The weather may or may not be drying up. We are, however, interested in whether precipication becomes more intense over time.

To see that, we will count the overall rainfall over the year and the number of rainy days in each year. Then, by dividing the rainfall by the number of rainy days we will get an indication of whether we are getting rain in more concentrated bursts. We will then create a plot showing the ratio of rainfall over rainy days over the years. On the plot we will overlay the 10 years rolling average.