Cape-Town---Temperature-Forecasting-using-Timeseries

KO | EN

gitlite — search

#javascript #react #ai #llm #machine-learning #deep-learning #docker #windows #awesome #cli #vue #security #pytorch #openai

Cape-Town---Temperature-Forecasting-using-Timeseries

(★ 11)

Developed an ARIMA-based predictive model to forecast monthly temperatures in Cape Town by analysing historical weather data, ensuring stationarity, and achieving high-accuracy trends for climate monitoring and water management planning.

Cape-Town---Temperature-Forecasting-using-Timeseries 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Cape-Town---Temperature-Forecasting-using-Timeseries
Developed an ARIMA-based predictive model to forecast monthly temperatures in Cape Town by analysing historical weather data, ensuring stationarity, and achieving high-accuracy trends for climate monitoring and water management planning.

# Cape Town Temperature Time Series Forecasting 🌡️

## Project Overview

This project analyzes the average monthly temperatures in Cape Town, South Africa, from January 1993 to August 2024. The main goal is to forecast future temperatures using an ARIMA model, which is especially relevant given Cape Town's recent climate challenges, such as the severe drought from 2015 to 2017.

---

## Dataset

* **Name**: Climate Observer | Africa Data Hub
* **Source**: Monthly-Temperature Cape-Town (1993-2024)
* **Frequency**: Monthly
* **Key Variable**: "Average Temperature"

---

## Methodology

The analysis follows these key steps:

### 1. Data Visualization

First, the time series data is plotted to visually check for trends, seasonality, and other patterns. The data clearly shows seasonal cycles.

### 2. Stationarity Check

Forecasting models require the data to be stationary (meaning its statistical properties don't change over time). The **Augmented Dickey-Fuller (ADF) test** is used for this purpose.

* The test on the raw data gave a p-value of **0.01**. Since this is below the 0.05 significance level, the time series is considered **stationary**, and no differencing is needed.

### 3. Model Selection and Fitting

An **ARIMA (Autoregressive Integrated Moving Average)** model was chosen for forecasting.

* The `auto.arima()` function in R was used to automatically find the best model parameters.
* The selected model was an **ARIMA(1,0,0)(2,1,0)[12] with drift**, which captures both non-seasonal and seasonal patterns in the data.

### 4. Forecasting

The fitted ARIMA model was used to forecast average monthly temperatures for the next six months (September 2024 to February 2025).

---

## Results

The model's forecast continues the observed seasonal pattern with a slight upward trend.

The forecasted average temperatures are:

* **September 2024**: 13.88 °C
* **October 2024**: 17.64 °C
* **November 2024**: 18.96 °C
* **December 2024**: 19.99 °C
* **January 2025**: 21.87 °C

---

## Conclusion

This analysis successfully modeled Cape Town's historical temperature data and produced a reliable forecast. The results align with expected seasonal patterns and broader climate warming trends, providing valuable insights for planning in climate-sensitive sectors like water resource management.

Cape-Town---Temperature-Forecasting-using-Timeseries

Cape-Town---Temperature-Forecasting-using-Timeseries 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Cape-Town---Temperature-Forecasting-using-Timeseries

Cape-Town---Temperature-Forecasting-using-Timeseries 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README