Develop a proxy model to predict enhancement of C in atmospheric CO_{2} due to fossil fuel emissions (C_{ff}) based on inexpensive measurements of anthropogenic trace gases. If successful, apply to NOAA tower network and aircraft measurements.

Since fossil fuels are devoid of ^{14}C, the measured ^{14}C ratio of atmospheric CO_{2} (expressed as Delta-C) is the gold standard measurement to determine the fossil fuel enhancement, C* _{ff}*, to any atmospheric CO

In the PPR model, the multivariate surrogate gases data vector is projected along particular direction vectors. Each direction vector is determined when fitting the PPR model to the data. The projection along each direction vector is smoothed. The form of the smoother is adjustable.We select the complexity and form of the Projection Pursuit Regression model with a statistical learning method called cross-validation. For each particular from of the PPR model, we fit the model to training data. We select the optimal model by minimizing the mean-square difference between observed and validation data. After selecting the model by this cross-validation procedure, we then predict C* _{ff}* values for test data excluded from the modeling building process.

- The performance of the PPR depends critically on which surrogate gases are included in the model and the complexity and form of the PPR model. We focused on a PPR model where the smoothing functions are regression splines with adjustable degrees-of-freedom. We jointly determined which gases to include in the model and the effective degrees-of-freedom of the smoothing splines by cross-validation.
- The root-mean-square difference between predicted and observed C
_{ff}for the test data is approximately 1.1*µ*mole/mole. Since the uncertainties of C_{ff}measurements due to random and systematic sources are approximately 1*µ*mole/mole and 0.5*µ*mole/mole, the PPR proxy model is a promising candidate to predict C_{ff},and expand the number of C_{ff }measurements throughout North America by a factor of 3 to 4. - Based on knowledge of the random uncertainty and systematic uncertainty of C
_{ff }measurements, we determined that the difference between the surrogate model prediction for C_{ff}and observed C_{ff}were not statistically significant. In the deviation (observed - predicted) time series, there may be some interesting structures. Some structures may arise since systematic errors may vary temporally and spatially. This topic is currently under study. - We studied alternative models including: a multivariate linear regression proxy model,a one-dimensional linear surrogate model, and a Multivariate Adaptive Regression Spline (MARS). According to cross-validation, the PPR model was superior to each of these alternative model.
- Kevin Coakley, John Miller, Scott Lehman, Steven Montzka, Arlyn Andrews, and Ben Miller, "Creating a mulit-gas proxy for Delta-14 atmospheric fossil fuel-CO2," presentation at the 2013 Fall Meeting of the American Geophysical Union.
- A draft manuscript to be submitted to a peer-reviewed journal is undergoing NIST review.

Created April 8, 2015, Updated May 2, 2016