Skip to main content
eScholarship
Open Access Publications from the University of California

Sources of methane and nitrous oxide in California's Central Valley estimated through direct airborne flux and positive matrix factorization source apportionment of ground-based and regional tall tower measurements

  • Author(s): Guha, Abhinav
  • Advisor(s): Goldstein, Allen H
  • et al.
Abstract

Methane (CH4) and nitrous oxide (N2O) are two major greenhouse gases that contribute significantly to the increase in anthropogenic radiative-forcing causing perturbations to the earth's climate system. In a watershed moment in the state's history of environmental leadership and commitment, California, in 2006, opted for sharp reductions in their greenhouse gas (GHG) emissions and adopted a long-term approach to address climate change that includes regulation of emissions from individual emitters and source categories. There are large CH4 and N2O emissions sources in the state, predominantly in the agricultural and waste management sector. While these two gases account for < 10% of total annual greenhouse gas emissions of the state, large uncertainties exist in their `bottom-up' accounting in the state GHG inventory. Additionally, an increasing number of `top-down' studies based on ambient observations point towards underestimation of their emissions in the inventory.

Three intensive field observation campaigns that were spatially and temporally diverse took place between 2010 and 2013 in the Central Valley of California where the largest known sources of CH4 and N2O (e.g. agricultural systems and dairies) and potentially significant CH4 sources (e.g. oil and gas extraction) are located. The CalNex (California Nexus - Research at the Nexus of Air Quality and Climate Change) field campaign during summer 2010 (May 15 - June 30) took place in the urban core of Bakersfield in the southern San Joaquin Valley, a city whose economy is built around agriculture and the oil and gas industry. During summer of 2011, airborne measurements were performed over a large spatial domain, all across and around the Central Valley as part of the CABERNET (California Airborne BVOC Emission Research in Natural Ecosystem Transects) study. Next, a one-year continuous field campaign (WGC 2012-13, June 2012 - August 2013) was conducted at the Walnut Grove tall tower near the Sacramento-San Joaquin River Delta in the Central Valley.

Through analysis of these field measurements, this dissertation presents the apportionment of observed CH4 and N2O concentration enhancements into major source categories along with direct emissions estimates from airborne observations. We perform high-precision measurements of greenhouse gases using gas analyzers based on absorption spectroscopy, and other source marker volatile organic compounds (VOCs) using state of the art VOC measurement systems (e.g. proton transfer reaction mass spectrometry). We combine these measurements with a statistical source apportionment technique called positive matrix factorization (PMF) to evaluate and investigate the major local sources of CH4 and N2O during CalNex and Walnut Grove campaigns. In the CABERNET study, we combine measurements with an airborne approach to a well-established micrometeorological technique (eddy-covariance method) to derive CH4 fluxes over different source regions in the Central Valley.

In the CalNex experiments, we demonstrate that dairy and livestock remains the largest source sector of non-CO2 greenhouse gases in the San Joaquin Valley contributing most of the CH4 and much of the measured N2O at Bakersfield. Agriculture is observed to provide another major source of N2O, while vehicle emissions are found to be an insignificant source of N2O, contrary to the current statewide greenhouse gas inventory which includes vehicles as a major source. Our PMF source apportionment also produces an evaporative/fugitive factor but its relative lack of CH4 contributions points to removal processes from vented emissions in the surrounding O&G industry and the overwhelming dominance of the dairy CH4 source.

In the CABERNET experiments, we report enhancements of CH4 from a number of sources spread across the spatial domain of the Central Valley that improves our understanding of their distribution and relative strengths. We observe large enhancements of CH4 mixing ratios over the dairy and feedlot intensive regions of Central Valley corresponding with significant flux estimates that are larger than CH4 emission rates reported in the greenhouse gas inventory. We find evidence of significant CH4 emissions from fugitive and/or vented sources and cogeneration plants in the oil and gas fields of Kern County, all of which are minor to insignificant CH4 sources in the current greenhouse gas inventory. The CABERNET campaign represents the first successful implementation of airborne eddy covariance technique for CH4 flux measurements.

At Walnut Grove, we demonstrate the seasonal and temporal dependence of CH4 and N2O sources in the Central Valley. Applying PMF analysis on seasonal GHG-VOC data sets, we again identify dairies and livestock as the dominant source of CH4. A clear temporal dependence of emissions originating from a wetlands / Delta CH4 source is observed while CH4 contributions are also observed from a source originating from upwind urban and natural gas extraction activities. The agricultural soil management source of N2O has a seasonal dependence coincident with the agricultural growing season (and hence, fertilizer use) accounting for a majority of the N2O enhancements during spring and summers but being reduced to a negligible source during late fall and winters when manure management N2O emissions from dairy and livestock dominate the relative distribution. N2O is absent from the `urban' source, in contrast to the significant contribution to the statewide N2O inventory from vehicle emissions.

The application of greenhouse gas source apportionment using VOC tracers as identification tools at two independent sites in the Central Valley over vastly different temporal resolutions provide significant insights into the regional distribution of major CH4 sources. Direct airborne eddy covariance measurements provide a unique opportunity to constrain CH4 emissions in the Central Valley over regional spatial scales that are not directly observable by ground-based methods. Airborne observations provide identification of `hotspots' and under-inventoried CH4 sources, while airborne eddy covariance enables quantification of emissions from those area sources that are largely composed of arbitrarily located minor point sources (e.g. dairies and oil fields).

The top-down analysis provides confirmation of the dominance of dairy and livestock source for methane emissions in California. Minor but significant contributions to methane emissions are observed from oil and gas extraction, rice cultivation and wetlands; the estimates for these sectors being either negligible (e.g. wetlands) or highly uncertain (e.g. oil and gas extraction) in the statewide inventories and probably underestimated as a proportion of the total inventory. The top-down analysis also confirms agricultural soil management and dairy and livestock as the two principal sources of N2O consistent with the inventory, but shows that N2O contributions attributed to the transportation sector are overestimated in the statewide inventory. These new top down constraints should be used to correct these errors in the current bottom-up inventory, which is a critical step for future assessments of the efficacy of emission reduction regulations. Particularly, measurement techniques like vehicle dynamometer emission calculations (for transportation sources), source-specific short range ground-based inverse dispersion (for dairy and livestock sources), airborne eddy covariance and airborne mass balance approach based emissions estimation (over oil and gas fields) and ground based eddy-covariance (for wetlands and agriculture sector) can be used effectively to generate direct emissions estimates for methane and nitrous oxide that help update and improve the accuracy of the state inventory.

Main Content
Current View