Fuel use for irrigation pumping was calculated as the product of the number of diesel‐ powered irrigation pumps in the county, the estimated annual activity of each pump , the average brake horsepower of pump engines in the Yolo/Solano Air Quality Management District, and the brake specific fuel consumption per hour . As of 2003, an estimated 643 diesel‐powered irrigation pumps were operated in Yolo County . Statewide, the number of diesel irrigation pumps was projected to increase by 3.5 percent between 1990 and 2010 . In this inventory we estimated the 1990 and 2008 pump populations based on an assumption that the statewide trend was proportional to the increase in the number of pumps in the county. Input values for engine activity and average engine horsepower were taken from recent government reports which inventory statewide emissions from diesel irrigation pumps . The amount of CO2, N2O, and CH4 emitted was the product of the total amount of diesel fuel consumed by pumps operated in the county and the emission factor for each gas . A GIS database of rice field locations, area‐weighted SSURGO14 soil data and daily weather data from CIMIS15 was compiled for 66 rice fields in Yolo County. The DNDC model was validated against field results from two California‐based field experiments which assessed the affects of residue and water management on CH4 emissions .
Results from the model runs allowed us to generate an emissions factor specific to each management scenario by averaging the simulated CH4 emissions rates across all 66 fields. To account for changes in practice over time,blueberry grow bag size we assumed that 100 percent of the harvested rice area in 1990 was managed according to Scenario A, while in 2008 rice area was divided equally between Scenarios B and C. The percentage of the county’s rice area attributed to each scenario is roughly consistent with statewide estimates for residue burning, residue incorporation, and winter flooding . Finally, the management‐specific emissions factors were multiplied by the area under each management scenario and then summed across each management category to give the total CH4 emissions from rice cultivation for a given year. Further details on residue inputs, fertilizer rates, water management, and model calibration can be found in Sumner et al. and Holst and Buttner . A California‐specific method developed by the CARB was used to estimate emissions from residue burning. This approach is based on studies conducted by the University of California at Davis, which established emissions factors for CO2, N2O, and CH4 for the most commonly burned residues in California . For each gas, the harvested area was multiplied by the fraction of area burned, the crop mass burned per unit area, one minus the residue moisture content, and the corresponding emissions factor for each crop .
In the case of rice, the fraction burned was assumed to have declined from 99 to 11 percent between 1990 and 2008, which is consistent with statewide trends . Carbon dioxide emissions from the addition of limestone and urea were determined by multiplying the amount of each material applied in Yolo County by its default emission factor . The amount of each material was based on county sales records . In Yolo County, total agricultural emissions declined by 10.4 percent between 1990 and 2008 . The primary reason for this generalized decline was a notable reduction in both direct and indirect N2O emissions . Direct N2O emissions were the largest source of emissions during both inventory years, but decreased by 23.1 percent over the study period due to a countywide reduction in the amount of synthetic N fertilizer applied . This reduction in fertilizer use was driven by two important land use trends: a 6 percent reduction in the county’s irrigated cropland ; and a general shift away from crops that have high N rates coupled with an expansion in alfalfa and grape area which require less fertilizer . The large expansion of alfalfa acreage resulted in a moderate increase in the direct N2O emissions from crop residues , but this increase was not enough to offset the overall savings achieved by the displacement of corn and tomatoes. The direct N2O emissions from urine in pasture and manure application ranged between 5 percent and 15 percent of the total direct emissions and showed a small rise over the study period due to a proportional increase in livestock population. Estimates of nitrate lost through leaching and runoff accounted for approximately two‐ thirds of the indirect N2O emissions countywide, with ammonia volatilization responsible for the remaining one‐third .
More than 90 percent of indirect emissions originated from synthetic N fertilizers, while urine and manure from livestock were relatively minor sources. Consequently, the notable decline in indirect N2O emissions was also due to a decrease in the amount of synthetic N applied countywide. In both years, emissions of CO2, N2O, and CH4 from diesel‐powered mobile farm equipment were responsible for 20.0 to 23.0 percent of total agricultural emissions in Yolo County . This category showed little change in emissions over time in 1990 and 69.0 kt CO2e in 2008. This was because an increase in fuel consumption per unit area for several important crops offset the small decline in irrigated cropland . Total emissions from mobile farm equipment were 4 percent lower using the Tier 1 method as compared to estimates generated using the OFFROAD model . However, since the OFFROAD model uses equipment population and hourly usage data to estimate emissions, results from this Tier 3 method could not be used to disaggregate emissions by specific crop category. In both years, emissions of CO2, N2O, and CH4 from diesel‐powered mobile farm equipment were responsible for 20.0 to 23.0 percent of total agricultural emissions in Yolo County . While a reduction in county’s irrigated cropland may have been expected to save fuel and reduce associated emissions, this category showed little change in emissions over time . This was because an increase in fuel consumption per unit area for several important crops offset the small decline in irrigated cropland . Total emissions from mobile farm equipment were 4 percent lower using the Tier 1 method as compared to estimates generated using the OFFROAD model . However, since the OFFROAD model uses equipment population and hourly usage data to estimate emissions,blueberry box results from this Tier 3 method could not be used to disaggregate emissions by specific crop category. Diesel‐powered irrigation pumps emitted approximately 39.6 kt of CO2e in 1990 and 41.0 kt of CO2e in 2008 . This was equal to 11.7 to 13.5 percent of the total agricultural emissions. While irrigated cropland in the county has decreased overall, the amount of land with access to groundwater has continued to expand as new wells are drilled. The small increase in the number of wells operating in the county, therefore accounts for the proportional rise in emissions from irrigation pumping.In Yolo County, CH4 emissions from livestock contributed between 7.8 and 10.5 percent of the total agricultural emissions depending on the inventory year . This is lower than the proportion attributed to livestock statewide, which was more than 50 percent of all agricultural emissions in 2008 . The lower figure for the county essentially reflects the small number of dairy farms operated locally. By contrast, enteric fermentation from pasture‐raised beef cattle was the largest source of CH4 emissions from livestock in both inventory years . Since beef cattle and sheep populations have changed little since 1990, emissions from these livestock types were also relatively stable. While dairy cattle represented only 5 to 12 percent of the county’s cattle in any given year, an increase in the number of dairy cattle from approximately 800 to 2300 animals over the study period resulted in a 20.0 percent increase in total CH4 emissions from livestock .
Using the Tier 1 method prescribed by CARB, emissions of CH4 from rice cultivation were estimated to increase from 25.9 to 31.2 kt CO2e between 1990 and 2008 . This increase was entirely due to a 20.3 percent expansion in the area under rice cultivation . Estimates generated using the DNDC model showed a larger increase in emissions over the study period ; this Tier 3 method accounted for changes in residue and water management in addition to the increase in cultivated area . Emissions of N2O and CH4 from residue burning contributed 2.0 percent to the total agricultural emissions in 1990 and declined to less than 1.0 percent in 2008, due to the phasing out of rice straw burning in accordance with State regulations . Emissions of N2O and CH4 were relatively small compared to the amount of CO2 emitted during combustion . Most inventory guidelines consider CO2 from residue burning to be a “bio-genic” emission, arguing that it is theoretically equivalent to the CO2 generated during the decomposition of the same crop residue in the soil over the course of the year . Consequently, CO2 from residue burning has been excluded from our inventory total. Emissions of CO2 from lime and urea application each contributed approximately 1 percent to the overall agricultural emissions, and both declined over the study period .One of the main findings of this study is that emissions from agriculture in Yolo County were already on the decline long before the implementation of recent mitigation policies. This trend is largely market‐driven, arising from broad economic factors that are prompting local farmers to shift more of their land to crops which happen to require less N fertilizer and diesel fuel. For instance, many local farmers point to the strong markets for wine grapes and alfalfa, which require fewer inputs as the main factor behind their recent local expansion . These Tier 1 methods do not fully capture the extent to which some growers are reducing fertilizer and fuel use in response to the rising cost and market volatility of inputs, rather than mitigation per se . It should be noted that interviews with Yolo County growers have documented numerous strategies to decrease energy use in cost‐effective ways, but they are often not yet integrated into the cost and return studies for Yolo County crop production .Another important factor contributing to the overall reduction in agricultural emissions was the 8,000 hectare decline in irrigated cropland. This loss of irrigated cropland raises two important questions. First, what type of land use is the cropland being displaced by? And second, how does the carbon footprint of other land uses compare to that of agriculture? Four countywide land‐use trends may explain the decline. Cropland could either be: left fallow, converted to non‐irrigated rangeland, restored to natural habitat, or developed for urban and industrial use. Shifting land use from irrigated cropland to fallow, rangeland, or natural habitat will generally reduce anthropogenic GHG emissions. The same cannot be said for cropland that is developed for urban uses. Urbanization accounted for the loss of about 6,500 acres of agricultural land between 1992 and 2008 . In 1990, emissions sources associated with Yolo County’s urban areas accounted for approximately 86 percent of the total GHG emissions countywide, while unincorporated areas supporting agriculture were responsible for the remaining 14 percent . If calculated on an area‐wide basis the county’s urban areas emitted approximately 152.0 t CO2e ha‐1 yr‐1 . By contrast, our inventory results indicate that in 1990 Yolo County’s irrigated cropland averaged 2.16 t CO2e ha‐1 yr‐1 and that livestock in rangelands emitted only 0.70 t CO2e ha‐1 yr‐1 . This 70‐fold difference in the annual rate of emissions between urbanized land and irrigated cropland suggests that land‐use policies, which protect existing farmland from urban development, are likely to help stabilize and or reduce future emissions, particularly if they are coupled with “smart growth” policies that prioritize urban infill over expansion .While avoided conversion of farmland will help curb emissions from urban sprawl, keeping farmland intact also affords numerous opportunities to mitigate emissions through changes in agricultural practice or by sequestering carbon in soils, perennial crops, or woody vegetation. In considering mitigation options, strategies should not hinder adaptation to climate change, as this could lead to loss of agricultural viability and potential urbanization, a much greater source of GHG emissions per acre.