GHCNV4_Analysis

About

This simple program downloads the latest GHCN V4 QCU (quality controlled unadjusted, or raw) monthly station data and generates an annual, global, land-only mean temperature anomaly series using gridded anomalies. The results are output to a csv file for further analysis. The program is written in Python and uses the Pandas library for data manipulation.

The program also downloads and cleans adjusted global land temperatue anomaly datasets from NASA GISS and the Climate Research Unit at the University of East Anglia, then generates a simple website to visualize the data using the chartjs library.

The program is designed to be run from the command line and is intended to be run on a regular basis to keep the data up to date. It is also intended to be run in your local environment.

Setup

clone the repository

git clone https://github.com/aljones1816/GHCNV4_Analysis.git

cd GHCNV4_Analysis

Set up and activate a virtual environment

python3 -m venv venv
source venv/bin/activate

Install the required dependencies

pip3 install -r requirements.txt

To run the program

python3 main.py

If you see something like the following in your terminal, then Flask is up and running!

 * Running on http://
 Press CTRL+C to quit

To access the dashboard

Once you confirm that the Flask server is running, open your browser and navigate to the following URL:

http://127.0.0.1:5000

GHCN Analysis Overview

This dataset is a global land average. The method I used is extremely simple (so I've called the dataset simpleTemp), and applies no adjustments or corrections to the raw data whatsoever. The basic outline is as follows:

• The stations are averaged into 5x5 grids. This is because some areas of the globe have a higher station density than others, so those areas will be unequally weighted if the stations are simply averaged individually.
• The grid boxes are spatially weighted, since boxes nearer the poles are smaller than boxes nearer the equator, and some boxes are partially covered by ocean (this is only meant to be land temperature).
• The series represents the average of the monthly anomalies. This is because stations can be moved over time, or come on and offline, so the anomalies prevents spurious trends from being introduced into the data.

Importantly, I have omitted any additional processing steps (adjustments), including:

• There is no correction made for time of observation bias. This produces a very small bias for the whole globe, but is significant for regional changes, particularly in the contiguous US, where the station network is particularly dense and transient.
• There is no homogenization applied. This means that urbanization bias and station moves are not accounted for. This, again, only produces a slight bias for the global mean trend, but can be significant at the regional level.
• There is no spatial interpolation for grid boxes that have no stations - these cells are simply absent from the dataset. As a result, the polar regions are under-sampled, and because the Arctic is one of the fastest warming regions on Earth, this produces a negative bias in the global trend. This approach is similar to that used by HadCRUT in versions below 5, and consequently my analysis is most similar to HadCRUT.