Climate Change: Using MACs to Quantify Solutions
A summary of my school project in the form research comparing multiple climate change solutions using MACs.
A Pressing Problem
Climate change is an important issue taking the toll over the last century. Ever since its discovery in the 20th century, people have increased awareness about the problem.
Credit: IPCC 2018
First some background: This chart shows the Global Mean Surface Temperature (GMST) - simply the average temperature on Earth's surface - up to the last few years, and several model scenarios. The temperature is measured relative to the 1850-1900 period.
As clearly indicated by the chart, surface temperatures have risen dramatically ever since the industrial revolution had begun. Most of the change, as shown by the red line, is caused by human activity, and much less is caused by natural phenomenon. Predictions have not shown considerable slowdowns even in the best case.
Even if everything is trying to show a "climate doom", it may be the absolute worst case. Even if warming happens, and it is, limimting the increase still has benefits. The IPCC analysis resulting in the aforementioned chart also mentions that current emissions alone would at most add 0.5 degrees Celsius of warming over 2-3 decades or even centuries. In other words, these 'worrying' emissions are actually relatively small in the long run.
Nevertheless, the effects on all scales will increase with emissions, so cutting every last one is still important - never use the above paragraph as an excuse for delays.
Research Purpose
This section is a short background, and is not rigorous. The next part defines the actual values and measures used
The research is conducted as a means to guide future efforts in managing climate change solutions. Any solution will work, but only some efficiently work. In an issue this pressing, no wastage should happen. This drives the research and is the underlying motive.
Units and Fundamental Concepts: CO2eq and MAC
As stated above, the research will compare solutions based on the efficiency. Here the only factor used is price, and other variables are yet to be included. Therefore the definition is (somewhat) more rigorously stated as the cost-effectiveness of the solutions. Additionally, the goal is to reduce warming effect, thus the complete definition is the cost-effectiveness of the solutions in cutting warming effects (not total emissions) by a certain amount.
To measure this, both the cost and warming effect is needed. Cost is as straightforward as the price tag of the device you're reading this on: it's just a currency and pretty much well-established at this point so inflation and conversions are the only weak links.
Warming effects are more complex. Climate change is affected by many gasses. Most of the time this is carbon dioxide (CO2), however much more powerful gasses exist. This menas that measuring amount is simply not an option.
In spite of this researchers have noticed the issue and devised CO2eq (also known as CO2e or fully spelled as CO2 equivalent or carbon dioxide equivalnt - any is valid), a measure ranking the quantity and effect of emissions. As expected, the same composition of emissions will have a CO2eq proportional to its amount, but also different compositions are accounted fairly based on the effect they impose, not only on their amount.
By definition, 1 unit of CO2eq is the effect of 1 unit of CO2 (unit can be any reasonable amount: g, kg, metric ton, and others).
After this comes the Marginal Abatement Cost (MAC). This can be seen as the 'unit price' to reduce emissions by a certain amount. The cost and amount can be in any unit - therefore it is similar to density, which can have many units (g/cm3 and kg/m3 are both valid densities, and $/kg and $/g are both valid MACs).
Costs are - as previously mentioned - trickiest during conversions and inflations. To curb the problem, all costs are converted to US Dollars and then accounted for inflation (from the year of the given currency to 2022).
WARNING: The above paragraph contains arbitrary choices and may or may not be reliable and/or accurate. Treat with caution!
Research Method
Due to limitations secondary data analysis is used by collecting and processing data from several sources. Note: Despite being included on the original research the sources are not given here due to the lack of formal need.
To account for (and measure) variability in the solutions' cost-effectiveness, the minimum and maximum MACs are used (thus being a range). If this is not possible (in other words, only one value is given) then the minimum and maximum are both set to the single value given.
For the secondary sources used, only publications focusing Asian countries are used. However, only public technologies are used, and only free open-access publications are used. Moreover, sources must be published no earlier than 2018, while all fullfilling sources are picked with no specific precedence.
NOTE: Searching for Sources
To use only trusted sources, I have used Google Scholar instead of normal Google. When searching, the top results are read first (as normal). However, this creates potential bias towards well-ranking results compared to low-ranking ones. Despite knowing it to be optimized for research, I have no information on how the search engine works and any possible biases. This may or may not be impactful, so do take good note of this and perform additional research on the topic if deemed neccessary.
Collected Data and Processing
The table of collected data follow:
| Technology / Solution | Min. MAC ($2022 / ton CO2eq) | Max. MAC ($2022 / ton CO2eq) |
|---|---|---|
| Behavioral energy efficiency | -215 | -215 |
| Corn starch ethanol (US) | -20 | 350 |
| Renewable Portfolio Standards | 0 | 215 |
| Reforestation | 1 | 11 |
| Wind energy subsidies | 2 | 294 |
| Clean Power Plan | 12 | 12 |
| Gasoline tax | 20 | 53 |
| Methane flaring regulation | 22 | 22 |
| Reducing federal coal leasing | 37 | 76 |
| CAFE Standards | 54 | 350 |
| Agricultural emissions policies | 56 | 73 |
| National Clean Energy Standard | 57 | 124 |
| Soil management | 64 | 64 |
| Livestock management policies | 80 | 80 |
| Concentrating solar power expansion (China & India) | 113 | 113 |
| Renewable fuel subsidies | 113 | 113 |
| Low carbon fuel standard | 113 | 3282 |
| Solar photovoltaics subsidies | 158 | 2376 |
| Biodiesel | 169 | 282 |
| Energy efficiency programs (China) | 282 | 339 |
| Cash for Clunkers | 305 | 475 |
| Weatherization assistance program | 396 | 396 |
| Dedicated battery electric vehicle subsidy | 396 | 724 |
| Solar energy (Thailand, 2015) | 59 | 189 |
| Wind power (Thailand, 2015) | 30 | 45 |
| Hydro energy (Thailand, 2015) | -172 | -70 |
| E-fuel (small-scale deployment, projected values until 2050) | 1022 | 1533 |
| E-fuel (large-scale deployment, projected values until 2050) | 25 | 344 |
| Low carbon concrete | 58 | 349 |
As seen, the range of values is considerably large, but contains negative values as well, thus a simple log scale is not feasible.
To solve this problem, a custom normalization method is used. This involves using a linear normalization to convert the values from 0 to 1, adding 1, and then taking the natural log of the value. Formally, with \(M_{min}\) being the minimum MAC, \(M_{max}\) being the maximum MAC, and \(M\) being the current MAC, the normalized value is:
\(\ln{(\frac{M - M_{min}}{M_{max} - M_{min}}+1)}\)
Formula: Normalization formula used.
WARNING: The above paragraph contains arbitrary choices and may or may not be reliable and/or accurate. Treat with caution!
Here both maximum and minimum MACs are processed altogether. As a result, the ranges are still valid after the process and no quirks will occur. Additionally, the natural log is called with a value between 1 and 2, thereby preventing undefined (or even negative) values from appearing. The absolute minimum is \(\ln{(1)}=0\) if \(M=M_{min}\). Note that this will make the minimum value invisible in the plot, but may be solvable by adding an offset and is related with plotting rather than data processing.
Final Results and Conclusion
For plotting a span chart is used. Some people mistake it for a bar chart due to the similarity. However, a span chart is ideal for ranges. A span chart's bars show ranges of values, and the height of the top and bottom sides show the actual values. Therefore the width of a bar corresponds to the range.
Due to the sheer amount of data shown, the chart is unable to label each value clearly. As a result an interactive plot is used. The final result is attached below:
Behavioral energy efficiency is the lowest, and may be very difficult to target. For all other visible bars, hovering/tapping on the blue part gives information on the name of the solution and the normalized values as used on the chart.
The main jaw-dropper, behavioral energy efficiency is the most effective solution. This should be clear, though. It costs nothing to start saving resources, and it helps in many ways. It makes economic sense. And environmental sense. And it reduces infrastructure load. This is the best for all: you don't need the lights on for a closed room/office anyway....
However, another observation is the massive variability some policy-based solutions have. This is clearly visible by the long bars.
My suggested explanation boils down to inconsistencies. Despite having the same concept, different governments may implement the policies differently, and the response may also be varying. This causes unprecedented differences. Therefore, policies can be effective, but must be carefully considered based on the current situation, similar implementations, and the demogrraphy of the targeted area.
WARNING: The above paragraph contains unverified theories and may or may not be reliable and/or accurate. Treat with caution!
So, the key takeaway for normal people is simple: every action taken counts! Nothing is too small to do, since if more and more people do it the collective results are massive and not negligible.
