Geo-Python 2025 final exercise (40 points)

The final exercise in the Geo-Python course involves calculating and plotting seasonal weather anomalies to see how temperatures have changed in different seasons over the past 100+ years. We will be using weather data from the Sodankylä weather station in northern Finland.

As with the earlier exercises in the course you should save your modifications (regularly) in your GitHub repository.

The final exercise is due by 23:59 on Monday, November 10th, 2025.

What is different?

• You will be working individually rather than with a partner on this exercise
• We will provide general instructions about what to do, but you will need to determine how you will approach this exercise
• You will have an empty Jupyter notebook and will be responsible for determining how you will use the Markdown and Python cells to perform your analysis
• AI-LLM-OK - Use of AI large language models, such as ChatGPT, is OK for this exercise. Details can be found under the heading "Use of AI-LLM tools in this exercise" in this document.
• The exercise is worth 40 points in total

The goal

Your task is to create a four-panel plot showing the seasonal temperature anomalies for winter, spring, summer, and autumn for the years 1909-2019. A template for how the final plot should look can be found below.

The data

We have provided a data file for you with daily temperature data from Sodankylä from January 1908 to October 2020. The data was downloaded from the NOAA Global Historical Climate Network database as a text file. The first five lines of the data file can be found below in order for you to see the data format.

STATION           STATION_NAME                                       DATE     TAVG     TMAX     TMIN     
----------------- -------------------------------------------------- -------- -------- -------- -------- 
GHCND:FI000007501                                   SODANKYLA-AWS-FI 19080101 -9999    2        -37      
GHCND:FI000007501                                   SODANKYLA-AWS-FI 19080102 -9999    6        -26      
GHCND:FI000007501                                   SODANKYLA-AWS-FI 19080103 -9999    7        -27  

As you can guess, the temperatures here are given in degrees Fahrenheit and the important columns for you are as follows:

• DATE: The date in 'YEARMODA' format, where 'YEAR' is the year using 4 digits, 'MO' is the two-digit month, and 'DA' is the day of the month.
• TAVG: The average daily temperature in Fahrenheit
• TMAX: The maximum daily temperature in Fahrenheit
• TMIN: The minimum daily temperature in Fahrenheit

Missing data are identified with -9999 and more information about the data can be found on the NOAA website.

Important note about the data!

You will notice that there are many missing values in the TAVG column. For days where the TAVG values are missing, you can calculate an estimate of the average daily temperature by averaging the TMAX and TMIN values. You may want to do this in a new column. Note don't replace the existing TAVG values with your estimates!

What to do

In contrast to earlier exercises in the course, in this exercise we are providing an empty notebook and only general instructions of how to complete the exercise. This is to help us see how much you have learned about programming in Python and to see how you would approach interacting with data using your new skills.

For this exercise you should:

1. Read in the provided data file using pandas and convert missing data to NA values.

2. Fill in the missing values in the TAVG column with your estimates of the average daily temperature as noted above in the data section. You can then drop any average daily temperatures that are still missing.

3. Define and use a function to convert temperatures in Fahrenheit to Celsius.

4. Calculate seasonal average temperatures for each season in every year (e.g., Winter 1909, Spring 1909, Summer 1909, ...)

   • The seasons should include the following months:

     • Winter: December, January, February
     • Spring: March, April, May
     • Summer: June, July, August
     • Autumn: September, October, November

5. Calculate seasonal average temperatures for the reference period 1951-1980 (e.g., 4 values in total, one for each season)

6. Calculate seasonal temperature anomalies for each year

7. Plot the data as shown in the example above

Note that we also hope to see you use Markdown cells to explain your data analysis, and code comments where needed to explain what your code does.

Use of AI-LLM tools in this exercise

The use of artificial intelligence large language model (AI-LLM) tools (e.g., ChatGPT) is allowed in this exercise. If you choose to use such tools, you must also:

1. Clearly acknowledge which tool(s) and versions of such tools were utilized to complete the exercise in a section of your notebook titled "Acknowledgements".
2. Describe how the tool(s) was/were used to complete the exercise in a section titled Use of AI tools with enough detail that your results could be reproduced. For example, this would include:
   • Explaining how the tools have been used in general terms, such as descibing whether they have been used to generate draft text explaining what you have done for the exercise or to produce code you have used.
   • Listing all prompts that were used to produce code as well as the code that was produced (only the final code is needed if multiple prompts were used) for any sections where code has been generated using an AI tool.
     • Please also note if the generated code was modified by you and how.

Failure to acknowledge or describe your use of AI tools in this exercise (if used) is considered cheating and will be handled accordingly. Also, please be aware that although an AI tool may have been used to produce parts of your exercise, you are responsible for the results that are produced and should still complete all of the tasks listed above in the section titled "What to do".

Grading

Grading criteria for the final exercise are available on the course website.

The total number of points you can receive is 40.