State Space Models

All state space models are written and estimated in the R programming language. The models are available here with instructions and R procedures for manipulating the models here here.

Saturday, September 3, 2016

Coal Will Make Reducing Energy Intensity Difficult


May 26, 2026 Scientist Ditch Scary Climate Scenario Goodbye RCP8.5 (see below). A Problem with RCP8.5 was the overly pessimistic projection for coal use and it's environmental impact.

The New York Times ran an article on Aug 30 titled The Challenge of Cutting Coal Dependence. The article focuses on the problems Germany, "a leader in the push against climate change", is having reducing dependence on coal production. Coal is Germany's main and dirtiest source of electricity generation. And, the same is true for many other countries, for example the US, China, and India. The problem is how to replace all the jobs that would be lost. No one has a practical answer. 


The graph above of coal consumption over time (from the NYT article) shows that Germany and the US have basically stabilized their coal consumption. The rest of the World, on the other hand, has not. My main question, when I read the article, was what can we expect from the future.

The graph at the beginning of this post is a forecast of coal production based on the WL20 model (the forecast assumes no policy intervention in the future). It shows that, with relatively narrow bootstrap 98% prediction intervals, coal production will not stabilize until well after the year 2100.
Compare that forecast to the one, also drawn from the WL20 model, of oil production. The model predicts that oil production has peaked, again with a high degree of confidence, and will decline for the foreseeable future (with or without policy intervention).

While commentators have been optimistic about the role reduced oil production will have in future carbon emissions, they have missed the major roadblock to reducing carbon intensity. Coal is a plentiful and easily obtained resource. Mining coal creates jobs. The NY Times article concludes with a quote from Craig Morris, an environmental blogger: "Several degrees of warming by 2100 may sound scary, but not nearly as much as long-term joblessness just a few years from now."


Notes

My Global Temperature Projections aren't as bad as RCP8.5 or other worst-case scenarios. The reason is that World System Collapse reduces the drivers of climate change in World System input models.

For more information about data sources and how the State Space models were constructed, see the Boiler Plate.



WL20 Measurement Model


The measurement model for the World System has three historical environmental controllers: W1=(Growth-LivingPlanet), W2=(LivingPlanet-TEMP) and W3=(P.Oil.-TotalFootprint).  Information about the indicators (LivingPlanet index, Global TEMP, TotalFootprint), etc. can be found in the Boiler Plate.

** Recent data from NASA's CRES (Clouds and Earth’s Radiant Energy System) satellite data also indicates that Earth's Albedo (ability to reflect sunlight and reduce warming) is decreasing!




RCP Projections


RCP2.6 to RCP8.5


Although there has been a lot of controversy surrounding RCP8.5 (Scientist Ditch Scary Climate Scenario), one point to make about the graphic above is that there is no Unlimited-Exponential-Growth-Forever scenario which seems to be what Climate Deniers predict will happen. Neoclassical Economists (see the DICE model) argue that growth will be limited when Technological change (Productivity) and Population growth reach a steady state (See the Solow-Swan Neoclassical Growth Model below).


Solow-Swan Growth Model





The Solow-Swan Neoclassical Growth Model (the basis for the DICE model) is one of the most important macro models in economics (along with the AS-AD Supply-Demand model). The directed graph above shows that the two exogenous causal variables are Population (N) and Technology (TECH). If either of these variables stop growing, the system will reach a Steady State Economy. It may be surprising to students of Economics to learn that, from a causality standpoint, Labor (L), Production (Q) and Capital Stock (K) are dependent variables and are not the primary drivers of growth in the  Solow-Swan Neoclassical Growth Model.

UN Population projections currently anticipate that Population will peak at 10.3 Billion people in 2084. In order to reach a Steady State Economy, the  DICE model has to also assume (arguably) that Technology Change will also peak at some time in the future.