Progress 03/15/14 to 03/14/16
Outputs
Target Audience:The primary target audience this reporting period was the earth sciences community at large, via the submission of a paper the Journal of Climate. Results of this research were also presented to the earth sciences community at Oregon State University (mainly faculty and graduate students) in a presentation at a univeristy seminar series. Changes/Problems:One additional question we hoped to address was how sensitive the attribution results are to particular choices of climate model parameters, given there is uncertainty in what the "best" parameter values are. Our intent was to carry out a "perturbed physics" experiment in which multiple parameters were changes (on other words, "perturbed") and the climate was simulated with actual GHG forcing and again with pre-industrial level forcing, resutling in many tens of thousands of runs. Initial simulations with alternative parameter sets resulted in unreastic global climates (e.g. very large-top-of-atmosphere energy imbalances and excessive rainfall in some tropical areas), so we deemed it inappropriate to use these in an attribution study. With hindsight, this was an overly ambitious objective given the modest funding level ($50K). However, this same research team is continuing research into parameter senstivity in the context of climate impacts on drought and subsequent vegetation die-off in a much larger USDA-NIFA funded project: EaSM2 Forest Die-off, Climate Change, and Human Intervention in Western North America (OREW-2013-00628, PI: Mote). Simulations currently being done under this project will allow us to explore the robustness of 2012 drought attribution results under parameter uncertainty. What opportunities for training and professional development has the project provided?Mentorship of a PhD candidate. The graduate student received training in experimental design, coupled global/regional climate modeling, climate model evaluation and climate data analysis. How have the results been disseminated to communities of interest?Results of the research have been disseminated in two journal articles and one presentation at a university seminar series. Rupp, D. E., S. Li, N. Massey, S. N. Sparrow, P. W. Mote, M. R. Allen. 2015. Anthropogenic influence on the changing likelihood of an exceptionally warm summer in Texas, 2011, Geophysical Research Letters (42), 2392-2400, doi: 10.1002/2014GL062683. Rupp D. E., S. Li, P. W. Mote, N. Massey, S. Sparrow, D. C. H. Wallom. Influence of the ocean and greenhouse gases on severe drought likelihood in the central US in 2012. Journal of Climate. In review. Rupp, D. E. 2015. Strength in numbers: Investigating the impact of anthropogenic greenhouse gases on regional climate and extreme events using very large ensembles from a climate model. Physics of Oceans and Atmospheres Seminar Series, Oregon State University, Corvallis, OR, 24 Nov. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Project impact The US experienced a severe drought in 2012, comparable with the worst 1930's Dust Bowl droughts. There is concern that such once-rare events may already be more likely due to influence of rising greenhouse gases (GHGs). If true, then climate change becomes more portentous than a benign gradual increase in temperature. To quantify the anthropogenic impact on drought frequency to date, we asked: What was the contribution of sea surface temperatures and GHGs to the 2012 drought? The question was answered probabilistically with the use of thousands of climate simulations under both actual 2012, and hypothetical "natural-only", conditions. The key conclusion is that by 2012, GHG concentrations in the atmosphere were high enough to affect the likelihood of an extreme drought - a drought as severe as the one that occurred in 2012. For example, simulations indicate that a summer rainfall deficit that would have occurred only every 100 years, on average, would occur every 30 to 50 years, given the state of the climate in 2012. The economic implications are substantial: What would it mean to the US agricultural sector, and those that depend on it, to take another US$30 billion hit during the next few decades? Goal: Determine if human activity increased the probability of the 2012 US Drought. Objectives: 1. Estimate the change in the probability distribution of extreme drought due to increase in greenhouse gases concentrations by the year 2012. 1.1. Major activities completed: The influence of anthropogenic GHGs in inducing the particular atmospheric and land surface conditions (in other words, soil moisture deficits) that lead to severe drought in the central US was estimated. To quantify this influence, a large number of simulations of the climate in 2012 were runs assuming 1) GHG concentrations at observed levels, and 2) GHGs concentrations at pre-industrial levels. In total, about 5,000 simulations were generated to examine the difference in climate under these two scenarios. 1.2. Data collected/produced: HadAM3P global climate model output given observed greenhouse gas concentrations for the years Dec. 2009 to Nov. 2012: The data consist of monthly atmospheric and surface variables from a large initial conditions ensemble of over 800 simulations per year. Used as the baseline for comparison against climate model simulations of a pre-industrial world. HadAM3P global climate model output given pre-industrial greenhouse gas forcing for the years Dec. 2009 to Nov. 2012: The data consist of monthly atmospheric and surface variables from a large initial conditions and sea surface temperature (SST) ensemble of over 4,000 simulations per year. Used to quantify the effect of anthropogenic greenhouse gases on US temperature, precipitation and soil moisture during the years 2011 and 2012. 1.3. Summary statistics/results: In the simulated environment, anthropogenic GHGs increase the odds of severe summer rainfall deficits, such as those that occur every 100 years on average, by a factor of 2 to 3. Viewed another way, anthropogenic GHGs result in about a 20% reduction in the rainfall amount that would have had a 100-year return period under pre-industrial conditions. The simulations show that anthropogenic GHGs increases do not change the likelihood of extremely dry soils. This is happens because during severe summer drought the decreases in actual evapotranspiration (despite the large evaporative demand) are of equal magnitude as the decreases in rainfall, so their changes effectively cancel each other. It is important to note that the climate model does not simulate irrigation, so these soil moisture responses results may not pertain to irrigated lands. 1.4. Key outcomes: The key outcome is the result the by 2012, GHG concentrations in atmosphere were high enough to affect the likelihood of an extreme drought - a drought as severe as the one that occurred in 2012. Simulations indicate that a precipitation deficit with a 100-year return period in the pre-industrial error would occur every 30 to 50 years, on average, given the state of the climate in 2012. This result runs contrary to a previous study that concluded that GHGs played no significant role in the rainfall deficit that occurred in 2012. There are two important implications to these results. The first is that it would be unwise to assume that severe droughts will not be more likely during the coming decades. The second is that, based on the contrasting results of two studies (each using different climate models), their still considerable uncertainty about the strength of the regional hydroclimatic response to anthropogenic GHGs. 2. Examine the particular atmospheric and land surface conditions that lead up to such a drought and permit its persistence. 2.1. Major activities completed: The influence of the ocean in inducing the particular atmospheric and land surface conditions (in other words, soil moisture deficits) that lead to severe drought in the central US was examined. To quantify this influence, a large number of simulations (100 runs per year) of the climate were made using a global climate model for the years 1986-2014. The simulations served as the baseline against with to compare particular atmospheric patterns and soil moisture conditions simulated for the year 2012. 2.2. Data collected/produced: HadAM3P global climate model output given observed greenhouse gas concentrations for the period Dec. 1985 to Nov. 2014: The data consist of monthly atmospheric and surface variables from initial conditions ensemble of 100 simulations per year. Used as the baseline for comparing the weather patterns of given year (e.g. 2012) against a long-term average. 2.3. Summary statistics/results: Simulations support the notion that higher-than-average spring temperatures initiated a "heat wave flash drought" that drove up evaporative demand and reduced soil moisture. The state of the ocean played a significant role in creating a warm spring, ultimately contributing to about 30% of the March-May soil moisture deficit. Soil moisture stayed at low levels through summer because of severe rainfall deficit. The immediate cause of the low rainfall was a series of high pressure patterns over the central US that suppressed convection and diverted weather systems north through Canada instead of across the US. Sea surface temperature patterns created conditions favorable to the formation of the high pressure over the central US; simulations indicate the ocean contributed to about 30% of the rainfall deficit in May-August; the other 70% was due to atmospheric circulation patterns that developed independent of ocean influence. 2.4. Key outcomes: The key outcome is the indication the ocean played a substantially larger role (~30% contribution) in bringing on the 2012 drought than previous studies have concluded. The implication is that observations of winter/spring, along with forecasts of summer sea surface temperatures, could be used to predict with a few months lead time if a severe drought like the one 2012 would be more likely in a given year.
Publications
- Type:
Journal Articles
Status:
Under Review
Year Published:
2016
Citation:
Rupp D. E., S. Li, P. W. Mote, N. Massey, S. Sparrow, D. C. H. Wallom. Influence of the ocean and greenhouse gases on severe drought likelihood in the central US in 2012. Journal of Climate. In review.
- Type:
Other
Status:
Other
Year Published:
2015
Citation:
Rupp, D. E. 2015. Strength in numbers: Investigating the impact of anthropogenic greenhouse gases on regional climate and extreme events using very large ensembles from a climate model. Invited talk, Physics of Oceans and Atmospheres Seminar Series, Oregon State University, Corvallis, OR, 24 Nov.
|
Progress 03/15/14 to 03/14/15
Outputs
Target Audience:The primary target this reporting period is the scientific community via the following publication in the journal Geophysical Research Letters. Rupp, D. E., S. Li, N. Massey, S. N. Sparrow, P. W. Mote, M. R. Allen. 2015. Anthropogenic influence on the changing likelihood of an exceptionally warm summer in Texas, 2011, Geophysical Research Letters (42), 2392-2400, doi: 10.1002/2014GL062683. Changes/Problems:One part of this study was to examine the sensitivity of the attribution results to particular choices of climate model parameters, given there is always some uncertaintly about any given model parameter value. Errors (human caused) in configuring the model runs using alternative parameters were made, resulting in some instances of the model crashing. The fact that this occured during the holiday season when both Oregon State University and our partners at Oxford University were often away, led to delays and the request for a no-cost extension. Problems were trouble-shot and runs successfully commenced in the spring of this year. We now have a large set model runs using alternative parameter sets and they are currently being analyzed. What opportunities for training and professional development has the project provided?Mentorship of a PhD candidate occurred during the summer 2014. The student received training in experimental design, coupled global/regional climate modeling, climate model evaluation and climate data analysis. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals? We will complete and submit at least one manuscript for a peer-reviewed scientific journal. We will present findings at a scientific conference. We will write, with assistance from a communications specialist, a brief article, accessible to a general audience, that can be circulated by (social) media outlets.
Impacts
What was accomplished under these goals?
Though the project is not complete, a preliminary analysis of global climate model output indicates that in 2012 anthropogenic GHGs slightly increased the probability of what in the pre-industrial era would have been a 1-in-50 year deficit in precipitation over the central US. Briefly, such "precipitation" droughts were made at least 2 times more likely in 2012 (best estimate is 3 times). One could also rephrase the question as this: How much dryer is the 50-year return period precipitation deficit? The modeling shows about a 5% reduction in May-August precipitation over the pre-industrial era at the 50-year return period. We can compare this to observed precipitation in the central US, which was 45% of normal in 2012. While direct comparisons of simulated modeled and observed precipitation need to be done with care, it provides a sense of the amount of precipitation deficit arising from GHG increases that were compounded with the precipitation deficit due to natural variability alone. The implication of this result is that the impacts of increasing GHGs are already being felt in the regional hydrological cycle despite large natural variabilty. Goals: To determine if human activity increased the probability of the 2012 US Drought. Objectives 1. estimate the change in the probability distribution of extreme drought due to increase in greenhouse gases concentrations by the year 2012. 1) Major activities completed / experiments conducted; A global climate model (GCM) was run under both a) observed 2011 and 2012 GHG forcing and observed sea surface temperatures (SSTs) and sea ice cover, and b) GHGs, SSTs and sea ice representing pre-industrial conditions. The GCM was run a very large number (thousands) of times with slight variations in the initial state of the atmosphere. Each of the simulations resulted in an alternate, plausible, series of weather during the course of a year. Some of these simulated generate severe droughts and heatwaves, others did not. The likelihood of severe modeled droughts was quantified with and without anthropogenic GHG concentrations at 2012 levels. Becuase we also simulated 2011 conditions leading up to 2012, we took the opportunity to explore the Texas heatwave and drought of 2011. 2) Data collected; No new data were "collected" for this project, other than output from the climate model runs. 3) Summary statistics and discussion of results and The analysis for 2012 is not complete, so we provide only preliminary results at this time. As stated above, in our simulated environment the 2012-level anthropogenic GHG concencentrations slightly increased the probability of what in the pre-industrial era would have been a 1-in-50 year deficit in precipitation over the central US during May-August period, which is when the observed drought was most intense. Quantified, such precipitation deficits were made 2 to 4 times more likely in 2012. Viewed in terms of change in precipitation deficit for a 1-in-50 year drought, it equates to approximately 5% less precipitation under 2012-level GHG concencentrations. A drought may be defined but measures other than precipitation deficit, so we are also examining soil moisture. Interestingly, the simulations show no detectable difference in summer soil moisture with and without anthropogenic GHGs. This may be because the increased winter and spring precipitation under 2012-level GHGs help maintain higher soil moistures at least into the earlier months of summer. We are still investigating the cause(s). 4) Key outcomes or other accomplishments realized. While the tendency toward decreased summer precipation under 2012-level GHGs is an important result, a key outcome may be the unexpected lack of change in likelihood of severe soil moisture deficitts despite the decrease in summer preciptiation and the expected increase in potential evapotranspiration (PET), the latter driven by higher regional temperatures. What appears to occur is that during severe drought, actual evapotranspiration (ET) simply tracks precipitation, even though PET is higher, such that soil moisture remains same. In effect, nder these conditions the system is nearly completely water limited, not energy limited, so the additional increase in temperature has very little effect on the net result. We should be careful about interpreting soil moisture results, however, because of simplifications in the land surface model; for example, vegetation is not dynamic throughout the growing season, vegetation is assumed to be unaffected by changing CO2 concentrations, and irrigation is not simulated. In other words, the lack of change in soil moisture does not necessarily imply a lack of change in stress on plants. Our analysis of the Texas 2011 heatwave show a sizable change in likelihood (~ times 10) of extreme high temperature anomalies due to anthropogenic GHGs (Rupp et al. 2011). An infuence on extreme precipitation deficits was not detected however, despite the reductions in mean summer precipitation arising frmo increasedl GHG concentrations. 2. examine the particular atmospheric and land surface conditions that lead up to such a drought and permit its persistence. 1) Major activities completed / experiments conducted; A major activity in support of Objective 2 was to carry out the modeling experiments as decribed for the under Objective 1. Further work towards this objective is stil being done. We note that Hoerling et al. (2014) have published on the proximate causes of the 2012 drought and the influence of SSTs. We will aim to not simply repeat their analysis, but provide additional insight made possible by our very large ensemble of simulations. 2) Data collected; No new data were collected for this project. 3) Summary statistics and discussion of results and Nothing yet to report. 4) Key outcomes or other accomplishments realized. Nothing yet to report. References Hoerling, M., J. Eischeid, A. Kumar, R. Leung, A. Mariotti, K. Mo, S. Schubert, and R. Seager, 2014: Causes and predictability of the 2012 Great Plains drought. Bull. Amer. Meteor. Soc., 95, 269-282. doi: 10.1175/BAMS-D-13-00055.1 Rupp, D. E., S. Li, N. Massey, S. N. Sparrow, P. W. Mote, and M. R. Allen (2015), Anthropogenic influence on the changing likelihood of an exceptionally warm summer in Texas, 2011, Geophys. Res. Lett., 42, 2392-2400. doi: 10.1002/2014GL062683.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2015
Citation:
Rupp, D. E., S. Li, N. Massey, S. N. Sparrow, P. W. Mote, M. R. Allen. 2015. Anthropogenic influence on the changing likelihood of an exceptionally warm summer in Texas, 2011, Geophysical Research Letters (42), 2392-2400, doi: 10.1002/2014GL062683.
|