Climate Research Project: Weather Forecast Accuracy for Pune, Maharashtra (November Study)

Are online weather predictions really accurate for your city? If you live in Maharashtra and especially in Pune, this is a very important question. Farmers, students, office workers and event planners all depend on weather forecasts to plan their daily activities. In this detailed guide, we will learn how to create a 30‑day climate research project that compares weather predictions from different agencies with the actual weather in Pune during the month of November.

This post is written in a simple way so that school and college students can use it directly as a project or assignment and adults can use it to understand how reliable online weather forecasts really are. We will compare the 5‑day‑ahead predictions from:

• Google Weather
• IMD (India Meteorological Department)
• ISRO‑based services such as Meghdoot
• Any one more private weather website or app (AccuWeather, Weather.com, Skymet, etc.)

The goal is to find out: Which source gives the most accurate 5‑day forecast for Pune in November?

Why Choose Pune and the Month of November?

Pune is one of the fastest growing cities in India, with a mix of urban and rural areas around it. Its weather in November is usually:

• Pleasant days – temperatures often between 27 °C and 32 °C
• Cool nights and early mornings – sometimes below 15 °C
• Post‑monsoon season – rainfall is normally very low but not impossible

Because the monsoon is almost over and winter is just starting, November is a good month to test the temperature accuracy of different weather prediction websites. The weather is not extremely unstable, so we can clearly see small differences between predictions and actual values.

Project Objective

The main objective of this climate research project is:

1. To collect 5‑day‑ahead weather forecasts for all 30 days of November for Pune, Maharashtra.
2. To record the actual weather (especially maximum temperature; you can also add minimum temperature and rainfall).
3. To calculate how much each forecast differed from the actual weather.
4. To compare Google, IMD, ISRO and another weather agency and decide which one is most accurate for Pune in November.

Materials and Data Sources

You do not need any expensive equipment to perform this study. A smartphone or computer with internet access is enough. Below are the recommended sources and how to use them.

1. Google Weather

Simply search for "Pune weather" on Google. You will see a small weather card with:

• Temperature for today
• Hourly values
• A 10‑day forecast

From this, write down the predicted maximum temperature for the date that is exactly 5 days ahead.

2. IMD (India Meteorological Department)

You can use the official MAUSAM app or the website mausam.imd.gov.in. Choose “City Forecast” and select Pune. Note the 5‑day forecast values, especially the expected maximum temperature for the desired date. IMD is the government authority for weather, so it is a very important reference.

3. ISRO‑Based Forecast (Meghdoot or Other)

ISRO works with IMD and ICAR to provide agro‑meteorological information, especially through the Meghdoot mobile app. Inside Meghdoot, set:

• State: Maharashtra
• District: Pune

You will see weather predictions that are mainly designed for farmers. Record the predicted maximum temperature for the date that is 5 days in the future.

4. Another Weather Agency (Your Choice)

To make your study complete, select one more popular service, such as:

• AccuWeather
• Weather.com
• Skymet Weather
• Windy, etc.

Open the website or app, search for Pune, Maharashtra, and use its multi‑day forecast to record the 5‑day‑ahead maximum temperature. Mention the name of this agency clearly in your project report.

Step‑by‑Step Methodology for 30 Days

To keep the project scientifically correct, we must be strict about the timing of data collection. Every forecast that you record must be taken exactly 5 days before the date you are studying. Here is the complete step‑by‑step method for November.

1. Plan Your Calendar

You want actual weather for 1 November to 30 November. Therefore you must start collecting 5‑day‑ahead forecasts from the last week of October:

• For 1 November, record forecasts on 27 October
• For 2 November, record forecasts on 28 October
• For 3 November, record forecasts on 29 October
• Continue this pattern until the end of November

2. Create a Data Table

You can draw this table in a notebook or create it in Excel / Google Sheets for easier calculation. Below is an example table structure for your project.

Date (Nov)	Google Max (°C)	IMD Max (°C)	ISRO Max (°C)	Other Agency Max (°C)	Actual Max (°C)	Average Difference to Actual (°C)
1
2
3
4
5
...
30

You can also add extra columns for minimum temperature and rainfall, but for most school projects, using only maximum temperature is enough and keeps the calculations simple.

3. Daily Recording Procedure

On each “forecast collection” day (for example, 27 October for 1 November), do the following:

1. Open Google, IMD, ISRO Meghdoot and your chosen other weather site or app.
2. For each one, find the forecast for the date that is exactly 5 days in the future.
3. Write down the predicted maximum temperature in your table.

On the actual date (for example, on 1 November itself), do this:

1. Go to IMD or any reliable “Past Weather” / “History” page for Pune.
2. Note the observed maximum temperature for that date.
3. Fill it in the “Actual Max” column of your table.

Repeat the same process for all 30 days of November.

4. Calculating the Forecast Error

To find out how accurate each forecast is, calculate the error for every day. For a particular date, suppose the actual maximum temperature is A and the predicted maximum temperatures are:

• G = Google
• I = IMD
• S = ISRO
• O = Other agency

Then the absolute error for each agency is:

• Error(Google) = |G − A|
• Error(IMD) = |I − A|
• Error(ISRO) = |S − A|
• Error(Other) = |O − A|

The vertical bars | | mean you ignore whether the difference is positive or negative; you only look at the size of the difference.

To get the average difference for that day (the value to fill in the last column of the table), use:

Average Difference = ( Error(Google) + Error(IMD) + Error(ISRO) + Error(Other) ) / 4

If you are using Excel or Google Sheets and your row looks like this:

• Column A – Date
• Column B – Google Max
• Column C – IMD Max
• Column D – ISRO Max
• Column E – Other Agency Max
• Column F – Actual Max
• Column G – Average Difference

then you can type this formula in cell G2:

=AVERAGE(ABS(B2-F2), ABS(C2-F2), ABS(D2-F2), ABS(E2-F2))

Copy this formula down for all 30 rows. The spreadsheet will automatically calculate the average forecast error for each day in November.

How to Analyse Your Results

Once your table is complete, you can perform a simple but powerful analysis that will look impressive in any school or college project report.

1. Overall Accuracy of Each Agency

Calculate the average error for each agency across all 30 days. Again, if you are using Excel or Google Sheets, and your data runs from row 2 to row 31, you can use formulas like:

• Google overall error:
  =AVERAGE(ABS(B2:B31 - F2:F31))
• IMD overall error:
  =AVERAGE(ABS(C2:C31 - F2:F31))
• ISRO overall error:
  =AVERAGE(ABS(D2:D31 - F2:F31))
• Other agency overall error:
  =AVERAGE(ABS(E2:E31 - F2:F31))

(If your spreadsheet software does not allow using ABS on entire ranges like this, you can instead create a separate “Error” column for each agency and then take the average of that column.)

The agency with the smallest average error is the most accurate at predicting the 5‑day‑ahead maximum temperature for Pune in November in your study.

2. Graphs for Better Presentation

Graphs make your climate research report look professional and are easy to understand. You can create:

• Line graph of actual vs predicted temperatures – Put dates on the x‑axis and temperature on the y‑axis. Draw one line for Actual, and one line for each agency.
• Bar graph of average error – One bar each for Google, IMD, ISRO and the other agency, showing their overall average error.

Most spreadsheet tools allow you to select your data and click “Insert Chart” to automatically create these graphs.

3. Discussion and Interpretation

In your written report, you can discuss points such as:

• Which agency was most accurate for November in Pune?
• Were the predictions better on clear days compared to days with clouds or rain?
• Did any agency consistently predict higher or lower temperatures than the actual values?
• What does this tell us about the reliability of online weather apps?

Remember that your results are for one city (Pune) and one month (November). Accuracy may be different in other months or places. Still, this project gives an excellent real‑life example of how data and statistics are used in climate science.

Suggested Structure for Your Final Climate Research Report

When you write the final report or make a project file, you can follow this simple structure:

1. Title Page
   Climate Research Project – Weather Forecast Accuracy for Pune, Maharashtra (November Study)
2. Acknowledgement
   Thank your teacher, parents, and any websites or apps used.
3. Introduction
   Importance of weather prediction, short description of Pune’s climate.
4. Objectives
   Clear bullet points about what you wanted to find out.
5. Data Sources and Tools
   Google Weather, IMD, ISRO, other agency, Excel / notebook, etc.
6. Methodology
   Step‑by‑step explanation of how you collected forecasts 5 days in advance and how you collected actual data.
7. Data Table
   Paste or attach the 30‑day table with all numbers.
8. Calculations and Graphs
   Show formulas and include graphs if possible.
9. Results and Discussion
   Explain which agency was most accurate and how big the typical errors were.
10. Conclusion
    Short summary of what you learned about weather prediction in Pune.
11. References
    List of websites and apps used in the project.

Conclusion

This climate research project on weather forecast accuracy for Pune, Maharashtra in November is practical, data‑driven and easy to understand. By following the steps explained above, you can create a strong project that is useful not only for school marks but also for your own understanding of how reliable different weather services are.

Such projects encourage students to think scientifically, question the information they see on their phones and analyse real observations from their own city. You can extend the same method to other months, other cities, or even compare short‑range and long‑range forecasts.

If you repeat this study every year, you will slowly build a powerful personal database that documents how climate and prediction accuracy are changing over time in Pune. This kind of citizen science is very valuable in the era of climate change.