Extreme Weather Events and Economic Losses: the Financial Benefit to Addressing Climate Change¶

Chapin Wilson - DS 2500 Project Proposal¶

Problem:¶

Many people view climate change mitigation in opposition to economic growth; reducing emmisions, promoting renewable energy, and investing in climate technologies are all costly and may inhibit industrialization and rapid innovation. However, recent data suggests that addressing global warming can actually protect economic prosperity.

Solution:¶

The goal of this project would be to research this claim and investigate the connection between one of the effects of the climate change (extreme weather) and gdp losses per country. I would like to see if a climate risk index rank that is particularly high or low corresponds with similarly high or low gdp losses. I would also like to cluster countries in a simiar region together and if they have similar rankings for both of these variables. If possible, I could create a classifier to predict which region a country is in based on these factors. This could also help identify regions that are especially in need assistance in terms of handling these extreme weather events and combatting climate change.

Impact:¶

Understanding the relationship between the climate and the economy could encourage specific countries to invest in climate change mitigation strategies which would help protect the planet and save people's lives.

The Dataset¶

A Kaggle Dataset with these attributes for almost every country (although some countries are missing certain values)

the_geom: Geometry of the country (Geometry)
the_geom: Geometry of the country (Geometry)
the_geom_webmercator: Web Mercator projection of the geometry of the country (Geometry)
cri_rank: Rank of the country on the Climate Risk Index (Integer)
cri_score: Score of the country on the Climate Risk Index (Integer)
fatalities_per_100k_rank: Rank of the country in terms of fatalities per 100,000 people (Integer)
fatalities_per_100k_total: Total number fatalities per 100,000 people (Integer)
fatalities_rank: Rank of the country in terms of total fatalities (Integer)
fatalities_total: Total number of fatalities (Integer)
losses_per_gdp__rank: Rank of the country in terms of losses per GDP (Integer)
losses_per_gdp__total: Total losses per GDP (Integer)
losses_usdm_ppp_rank: Rank of the country in terms of losses in USDM PPP (Integer)
losses_usdm_ppp_total: Total losses in USDM PPP (Integer)
rw_country_code: Country code (String)
rw_country_name: Country name (String)

In [9]:

from collections import defaultdict
import pandas as pd

cri_rank = defaultdict(lambda: 0)
gdp_loss_rank = defaultdict(lambda: 0)

with open("climate-risk-index.csv" , "r") as infile:
    for line in infile:
        line = line.strip().split(",")
        
        cri_rank[line[4]] = line[5]
        gdp_loss_rank[line[4]] = line[11]

df_cri_rank = pd.DataFrame(cri_rank, index = [0])
display(df_cri_rank)

df_gdp_loss_rank = pd.DataFrame(gdp_loss_rank, index = [0])
display(df_gdp_loss_rank)

	country	Saudi Arabia	Romania	Spain	Slovenia	South Sudan	Sierra Leone	South Africa	Serbia	Slovak Republic	...	Montenegro	Rwanda	Qatar	Puerto Rico	Samoa	Seychelles	Gambia	Togo	Trinidad and Tobago	Tonga
0	cri_rank	79	61	69	135	133	102	33	83	123	...	135	110	129	110	135	135	135	131	135	135

1 rows × 183 columns

	country	Saudi Arabia	Romania	Spain	Slovenia	South Sudan	Sierra Leone	South Africa	Serbia	Slovak Republic	...	Montenegro	Rwanda	Qatar	Puerto Rico	Samoa	Seychelles	Gambia	Togo	Trinidad and Tobago	Tonga
0	losses_per_gdp__rank	131	16	86	135	120	124	24	33	115	...	135	130	113	81	135	135	135	123	135	135

1 rows × 183 columns

These two data dictionaries show two specific features for each country: cri_rank (rank of each country on the Climate Risk Index) and losses_per_gdp_rank (rank of each country in terms of losses per gdp). While the other features of this dataset may be useful in this project, these are the two that I feel will be most helpful in this analysis.

Potential Problems:¶

The major problem I think I would have would be determing "regions" to place each country into. Using continents doesn't seem entirely appropriate because of the range of different numbers of countries in each continent. Some continents also have countries with similar weather patterns (i.e. Europe) while others have wildly diffent climates (i.e. North America). Clustering by region will be a challenge that will require some thought.

Method:¶

I would like to use a correlation matrix illustrate the relationship between climate risk and the economy. I will also implement a classifier (or potentially cluster countries together) to gain insight into regions that are especially vulernable (both in terms of being exposed to an extreme weather event, as well to high gdp losses)