Accepted Manuscript Exploring changes in rainfall intensity and seasonal variability in the southeastern U.S.: stakeholder engagement, observations, and adaptation Daniel R. Dourte, Clyde W. Fraisse, Wendy-Lin Bartels PII: DOI: Reference:
S2212-0963(15)00005-4 http://dx.doi.org/10.1016/j.crm.2015.02.001 CRM 32
To appear in:
Climate Risk Management
Please cite this article as: D.R. Dourte, C.W. Fraisse, W-L. Bartels, Exploring changes in rainfall intensity and seasonal variability in the southeastern U.S.: stakeholder engagement, observations, and adaptation, Climate Risk Management (2015), doi: http://dx.doi.org/10.1016/j.crm.2015.02.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Exploring changes in rainfall intensity and seasonal variability in the southeastern U.S.: stakeholder engagement, observations, and adaptation Daniel R. Dourte, Clyde W. Fraisse, Wendy-Lin Bartels Abstract The distribution of rainfall has major impacts in agriculture, affecting the soil, hydrology, and plant health in agricultural systems. The goal of this study was to test for recent changes in rainfall intensity and seasonal rainfall variability in the Southeastern U.S. by exploring the data collaboratively with agricultural stakeholders. Daily rainfall records from the Global Historical Climatology Network were used to analyze changes in rain intensity and seasonal rainfall variability. During the last 30 years (1985-2014), there has been a significant change (53% increase) in the number extreme rainfall days (> 152.4 mm/day) and there have been significant decreases in the number of moderate intensity (12.7 – 25.4 mm/day) and heavy (25.4 – 76.2 mm/day) rainfall days in the Southeastern U.S., when compared to the previous 30-year period (1955-1984). There have also been significant decreases in the return period of months in which greater than half of the monthly total rain occurred in a single day; this is an original, stakeholder-developed rainfall intensity metric. The variability in spring and summer rainfall increased during the last 30 years, but winter and fall showed less variability in seasonal totals in the last 30 years. In agricultural systems, rainfall is one of the leading factors affecting yield variability; so it can be expected that more variable rainfall and more intense rain events could bring new challenges to agricultural production. However, these changes can also present opportunities for producers who are taking measures to adjust management strategies to make their systems more resilient to increased rain intensity and variability. Keywords Climate change adaptation, rainfall intensity, stakeholder engagement, agriculture Author names, affiliations, contact Daniel R. Dourte, Agricultural and Biological Engineering, University of Florida, Corresponding author: ph-001.352.392.1864 ext.220; e-mail
[email protected] Clyde W. Fraisse, Agricultural and Biological Engineering, University of Florida, ph001.352.392.1864 ext.271; e-mail
[email protected] Wendy-Lin Bartels, Florida Climate Institute, University of Florida, ph-001.352.392.1864 ext.232; e-mail
[email protected]
1
1. Introduction 1.1. High-intensity rainfall: agricultural impacts and state of knowledge The intensity of rainfall has important impacts on the hydrology of a system, and these impacts can be very different at small and large spatial scales. The effects of rainfall intensity and variability are wide-ranging among different sectors and stakeholders. In agriculture, rainfall intensity affects the partitioning of rainfall into infiltration and runoff, impacting the movement of soil and the amount of water available in plant root zones to support crop growth (Morin and Benyamini 1977; Sharpley 1985; Langhans et al. 2010). Movement of nutrients and crop protection chemicals are also impacted by the timing and intensity of rainfall. Our changing climate has already presented a variety of challenges to agriculture (Walthall et al. 2012; Lobell et al. 2011). In particular, increased rainfall intensities may lead to increased soil loss, and many of the management alternatives for increased climate resilience in agriculture are those that reduce soil loss and improve soil health (Lal et al. 2012; Delgado et al. 2011; Howden et al. 2007). There is a growing body of evidence showing rainfall in some parts of the world has become characterized by increasingly common high-intensity rain events (Groisman et al. 2012; Higgins and Kousky 2013; Trenberth et al. 2003; Trenberth 2011; Dourte et al. 2012; Keggenhoff et al. 2014). Projections of future climate change also suggest increasingly intense rainfall in some locations (Sillmann et al. 2013; Mirhosseini et al. 2013), including the Southeastern U.S., but general circulation models (GCMs) have been shown to underestimate high intensity rainfall in the Southeast (DeAngelis et al. 2013). A study of rainfall intensity from 1958-2012 for the U.S. (Walsh et al. 2014, updated from Karl et al. 2009) found that for all regions of the contiguous U.S. there was an increase in the amount of precipitation falling in very heavy rain events (defined as the heaviest 1% of all daily rainfall amounts) from 1958 to 2012. The increase in the Southeast was 27% during the 1958-2012 period; this is based on the endpoints of a trendline fitted to the 55 years of annual depth of rainfall in the top 1% of daily amounts. Higgins and Kousky (2013) examined the percent changes in observed daily precipitation over the United States between two 30-year periods (1950–79 and 1980–2009). They observed increased u e s of hea ai fall ≥ days, but this was the highest intensity rainfall of their three fixed-threshold categories of daily rainfall. Seasonal analysis of return periods during 1980-2009 was completed for 10-year, 5-year, and 3-year rain events during 1950-1979. In much the Southeastern, U.S., they found decreased return periods for the 1980-2009 period across the three 1950-1979 return periods for JFM, JAS, and OND seasons. AMJ season showed somewhat longer return periods for the more recent 30-year period. However, the absence of fixed-thresholds in the study can make it difficult for some stakeholders to connect the measures of rainfall intensity to the possible impacts. A recent study of rainfall intensity 2
changes in the Central U.S. also compared adjacent time periods 1948-1978 and 1979-2009 (Groisman et al. 2012) and found that the biggest increases between the two 31-year periods were in the most extreme of their four fixed-threshold rain intensity categories (>154.9 mm). Several other demonstrations of increased rain intensity in the U.S., using a variety of rainfall intensity metrics and time periods (Kunkel et al. 2013), suggest that for the Eastern U.S. there is a robust trend of increasing occurrence of extreme rainfall. However, none of these studies explored the seasonality of changes in extreme rainfall, and there are no stakeholder-driven rain intensity thresholds, which can be important in order for climate science to be connected to management changes for reducing risk. 1.2. High-intensity rainfall: meteorological causes and attribution of change High-intensity rain events require sufficient atmospheric water and strong upward circulation which can be caused by a variety of different meteorological causes. In the Southeastern U.S., defined in a study by Kunkel et al. (2012) to include Alabama, Florida, Georgia, North Carolina, South Carolina, and Virginia (Kunkel et al. 2012), this combination of moisture convergence and upward motion is most commonly brought about by tropical cyclones (TC), extratropical cyclone near a front (FRT), extratropical cyclone near center of low (ETC), and mesoscale convective system (MCS). These four causes of extreme rainfall were attributed to 51% (TC), 34% (FRT), 7% (ETC), and 6% (MCS) of the 1-in-5 year rain events across the region from 1908 to 2009 (Kunkel et al. 2012). The meteorological causes of intense rainfall vary by location and by time of year; for example, in the Southeastern U.S. many more of the extreme rain events in the fall are caused by tropical cyclones than during other times of the year. The seasonal differences in meteorological causes can give some clues as to what might be driving a shift in frequency of intense rain events. There can be two distinct parts of an analysis of climate changes: detection and attribution. It is beyond the scope of this study to answer the question of attribution of the changes in rainfall characteristics, but we summarize two commonly cited contributors to rainfall changes: irrigation expansion and temperature increases. Irrigation has been shown to affect rainfall and runoff in parts of the U.S. Southwest (Lo and Famiglietti 2013), through the mechanism of increased evapotranspiration and water vapor export in Califo ia s Ce t al Valle . “i ila l , parts of the Southeastern U.S. were estimated to have increased summer precipitation, during the period from 1980-2000, compared to a counterfactual atmospheric general circulation model run without irrigation (Puma and Cook 2010). Some downwind regions of the Ogallala aquifer-irrigated parts of the Great Plains in the U.S. have been found to have increased rainfall that could be attributed to irrigation expansion (DeAngelis et al. 2010). Impacts of atmospheric temperature on increasing rain intensity have been suggested by several studies (Min et al. 2011; Trenberth 2011; Li et al. 2011; Wang et al. 2010). Increased temperature can lead to 3
increased atmospheric water (Trenberth et al. 2003), and the moisture convergence that occurs during storm development can result in higher-intensity rainfall than might be observed in a cooler atmosphere with correspondingly lower atmospheric water. 1.3. Stakeholder engagement and objectives In the Southeastern U.S., the importance of water stress has emerged repeatedly in discussions among agricultural stakeholders, including those who participate in the Row Crop Climate Learning Network (Bartels et al. 2012). This regional group of producers, Extension professionals, and scientists from Florida, Georgia and Alabama was initiated by researchers in the Southeast Climate Consortium (SECC) in 2010 as a space for knowledge exchange about climate-related risks (Bartels et al. 2012). Twice-annual workshops have sparked questions about appropriate adaptive water management practices, and producer-led fieldtrips have demonstrated possible strategies for coping with drought or extreme rainfall. Several p odu e s o side the sel es to e eathe at he s a d olle t thei o eather data. These learning network participants have reported that in recent years they have experienced more intense, less frequent rainfalls during the summer months. As a result, discussion emerged about what constitutes an extreme precipitation event in the region and what stakeholders might expect in the future. In this paper we present the results of a study of rainfall intensity and variability in the Southeastern U.S.; we tested for changes in rainfall using daily rainfall data from stations in Alabama, Florida, Georgia, North Carolina, and South Carolina. Our study adds several unique elements to the recent literature on observed changes in rainfall intensity and variability: 1) the seasonality of changes in fixed-threshold rainfall intensity is examined, 2) seasonal rainfall variability is examined using the same records and time periods used for analysis of rainfall intensity – this was expected to help clarify if rainfall intensity changes were connected to changes in seasonal rainfall variability, and 3) a stakeholder-developed rainfall intensity metric was analyzed here to provide a measure of rain intensity of particular interest to agricultural producers – this quantity was the return period of having at least half of the monthly rainfall occurring in a single day. Including stakeholders as participants in the research design and Extension of research can improve the emphasis on management solutions for climate risk management (Meinke et al. 2006; Howden et al. 2007; Howden et al. 2014). This helps overcome a common challenge in Extension of climate science, the sometimes weak connection between climate change science and management alternatives to reduce climate risks (Morris et al. 2014; Wojcik et al. 2014; Hill et al. 2014). The objective of this study was to determine if there have been changes in recent decades in the intensity and seasonal variability of rainfall in the Southeastern U.S. Additionally, it was our goal to demonstrate how engagement with agricultural stakeholders could improve the management-relevance of 4
climate research and could contribute to the introduction of a novel measure of rainfall intensity. 2. Methods 2.1. Fixed-thresholds approach for evaluating rainfall intensity changes Two common approaches used in examinations of historical changes in rainfall intensity are the fixed-threshold and percentile-defined threshold approaches (Groisman et al. 2012). In the fixed-threshold approach, the numbers of rain events exceeding a specified intensity value, or within a specified intensity range, are counted over a specified time period. The fixedthresholds can be defined based on impacts, stakeholder input, or any other criteria. Percentile-defined thresholds are developed separately for each location (station or grid cell or climate division) based on probabilities of rain intensities in the historical record. The advantage of the percentile-defined approach is that it accounts for the spatial differences in the characteristics of rainfall, but this can also be problematic as it can result in very different rainfall intensities between locations for a single chosen percentile threshold of rainfall intensity. This limits the information available for considering impacts of extreme rainfall because the 99th-percentile threshold for some location might be only a moderately intense rain event in terms of impacts, even though it is rare for that location; therefore, examining changes in that type of event might have less importance for understanding potential impacts than that of a 99th-percentile rain intensity in a location that results in an intensity threshold of large expected impact (Groisman et al. 2012). We elected to use the fixed-thresholds approach in this study because 1) it allows for simpler consideration of potential impacts across a region, and 2) the thresholds can be defined based on stakeholder input. The numbers of rainfall days within each category for the most recent 30-year period (19852014) were compared to those for the previous period (1955-1984) to test for changes in rainfall intensity in the four fixed-threshold categories. The time of year in which changes in rain intensity were observed was examined by counting the number of rainfall days in each rain intensity category for each of four seasons: December-January-February (DJF), March-April-May (MAM), June-July-August (JJA), and September-October-November (SON). Testing for statistically significant differences in numbers of fixed-threshold rain intensity was completed using the non-parametric Wilcoxon-Mann-Whitney test, comparing the counts for each of the four rainfall categories between the two 30-year periods of 1955-1984 and 1985-2014. The approach of comparing the statistical properties of adjacent, multi-decadal periods has been used in several recent studies of rainfall intensity and variability, and it has the advantage of reducing the sensitivity to small shifts in the time series end points that can impact trend-line fitting (Higgins and Kousky 2013, Groisman et al. 2012, Li et al. 2011, Wang et al. 2010). 5
While the primary aim of this work is to test for changes in rainfall intensity, we have included a discussion here about two possible causes of changes in rainfall intensity: temperature increases and irrigated area expansion. State-average temperature data were retrieved from the U.S. Climate Divisional Database (Guttman and Quayle, 1996) for 1955-2014. The irrigated cropland areas were retrieved from the twelve USDA Farm and Ranch Irrigation Surveys from 1954-2007 (USDA 2014). 2.2. Defining rainfall intensity thresholds with stakeholders Categories of thresholds for daily rainfall intensity were developed by combining stakeholder input and studies from recent literature (Groisman et al. 2012). The authors collaborated with agricultural stakeholders in the Row Crop Climate Learning Network (Bartels et al. 2013) in one of the g oup s t i e-annual meetings. An interactive exercise was designed to explore perceptions about rainfall intensity. Three participant groups (producers, Extension professionals, and researchers) were asked to determine quantitative thresholds that would match each of five rainfall categories (good, moderate, heavy, very heavy, and extreme). A portion of this discussion also focused on management solutions that can be used by producers to reduce risks from high-intensity rainfall. Following small-group discussions, participants compared results and considered implications for research and management (Bartels et al. 2013). The stakeholder-defined thresholds (Table 1) matched very closely to the four threshold-based categories used in recent research (Groisman et al. 2012). For the purpose of allowing for comparisons with other recent studies of rainfall intensity in other regions, the fixed-thresholds used to examine changes in rainfall intensities were: 1) moderately heavy precipitation (within a 12.7–25.4 mm/day range), 2) heavy precipitation (within 25.4–76.2 mm/day range), 3) very heavy precipitation (> 76.2 mm/day), and 4) extreme precipitation (> 152.4 mm/day). In addition to the four threshold-based categories of rainfall intensity used here, a unique measure of rainfall intensity was developed from input of agricultural producers. Some producers in the Southeastern U.S. inquired about how common it was for at least half of the monthly total rainfall to occur in a single day, and they were also interested in knowing if this quantity had been changing. Daily rainfall data were used at each combination of station/year/month to count the numbers days having at least half of the monthly total rainfall. Comparisons were made between the periods 1955-1984 and 1985-2014; Wilcoxon-MannWhitney tests were used to check for significant differences in occurrence of months in which a single-day rainfall provided greater than or equal to half of the monthly rainfall. 2.3. Seasonal variability in rainfall and description of data sources
6
Changes in seasonal variability in rainfall amounts were examined using daily data summed to produce seasonal totals. The frequency of exceeding the standard deviation in seasonal total rainfall was used to detect changes in variability in the last 30 years (1985-2014 compared to 1955-1984). Daily rainfall data from the Global Historical Climatology Network-Daily (GHCN-D; Menne et al. 2013) were used in this study. Stations were selected in five Southeastern U.S. states (AL, FL, GA, NC, SC) if they had observations beginning in 1955 or earlier and continuing through the end of 2014. Stations were excluded if there was a missing year of data during the 60-year period from 1955-2014. This excluded 156 stations, leaving a total of 259 stations for daily rainfall in the Southeastern U.S. State-average annual and seasonal temperature data were retrieved from the U.S. Climate Divisional Database (Guttman and Quayle, 1996) for 19542014. The irrigated cropland areas were retrieved from the thirteen USDA Farm and Ranch Irrigation Surveys from 1954-2013 (USDA, 2015). The El Niño Southern Oscillation (ENSO) phenomenon is a major driver of seasonal variability in rainfall in the Southeastern U.S.; we used the 3-month Oceanic Niño Index (ONI) from the Climate Prediction Center to classify the seasonal ENSO phases to examine for differences in ENSO episodes between 1985-2014 and 1955-1984. 3. Results 3.1. Daily threshold-based rain intensity changes A threshold-based approach was used to examine changes in rainfall intensities in four categories of daily rainfall intensity 1) moderately heavy precipitation (12.7–25.4 mm/day range), 2) heavy precipitation (25.4–76.2 mm/day range), 3) very heavy precipitation (> 76.2 mm/day), and 4) extreme precipitation (> 152.4 mm/day). As expected, there is considerable spatial variability in differences in rain intensity across the region; Figure 1 illustrates the station-specific changes in extreme daily rainfall between 1985-2014 and 1955-1984. The average increase in numbers of extreme rainfall days among the top 25% of stations that observed increases in extreme rain was 3.4 more extreme rainfall days, and the average change in numbers of extreme rainfall days among the bottom 25% of stations that observed decreases in extreme rain was -1.2. The per-station average increase in numbers of extreme rainfall days among all stations was 0.8 more extreme rainfall days in 1985-2014 compared to 1955-1984. The regionally-averaged changes in rainfall also show a tendency toward increasing extreme rainfall. There have been about 53% more extreme rainfall days during 1985-2014 compared to 1955-1984 (Figure 2); this increase in extreme rainfall between the two 30-year periods was statistically significant. There were 423 extreme rainfall days observed for the period from 1955-1984, but there were 649 of these days observed across the region from 1985-2014, a considerable increase in these potentially high-impact rain intensities. There were some
7
decreases in the numbers of moderate and heavy rainfall days (Figure 2), but these are generally the preferred o good categories of rainfall intensity for agriculture (Table 1). The greatest increases in extreme rainfall were observed in fall (SON – September-OctoberNovember) and summer (JJA – June-July-August), 151% and 48% increases, respectively (Figure 3). There were also significant increases (35%) in very heavy rain (> 76.2 mm/day) in the SON season. MAM season was observed to have significant decreases in numbers fixed-threshold rainfall across all four rain threshold categories (Figure 3), meaning there were significantly fewer rainfall days in MAM during 1985-2014 compared to 1955-1984. This is supported by a 15% reduction in region-wide MAM rainfall for the 1985-2014 period compared to 1955-1984 (283 mm and 334 mm, respectively). 3.2. Stakeholder-defined, monthly rain intensity metric The agricultural producer-defined rainfall intensity measure, the instance of having at least half of the total monthly rainfall in a single day (here abbreviated, HMD), is summarized in Figure 4. The regional average increase in HMD across all months was 17% for 1985-2014 compared to 1955-1984. All monthly HMD increases, with the exceptions of June, August, September, and December, were statistically significant. The average return period for occurrence of HMD, averaged across all months, is about once every 4.4 years for the period 1985-2014 (Figure 4); this was a 13% shorter return period than that of monthly-average HMD during the previous 30 years, 1955-1984. October had the shortest return period of HMD, 2.5 and 2.3 years for the 1955-1984 and 1985-2014 periods, respectively; July had the longest return period of HMD, 10.3 and 7.9 years for the 1955-1984 and 1985-2014 periods, respectively. Discussions with participants in the Row Crop Climate Learning Network also revealed that perceptions of e t e e ai e e ts a e shaped fa to s e o d just rainfall intensity. For instance, producers take into account the type of soil and crop, planting growth stage, as well as recent rainfall context. This emphasizes the importance of local context and interpretation when developing rain intensity thresholds. 3.3. Seasonal rainfall variability Seasonal rainfall totals show increased variability in spring (March-April-May) and summer (June-July-August) and decreased variability in fall (September-October-November) and winter (December-January-February) during the last 30 years (Figure 5). The increased variability in summer is consistent with what was shown by Li et al. 2011 and Wang et al. 2010, though their delineation of Southeastern U.S. was different than the five state region used here. The increasing frequency of extreme daily rainfall does not seem to be impacting seasonal rainfall variability, which is not surprising given the rarity of these rain intensities. Extreme rainfall represented 0.058% and 0.094% of all rainfall days (having greater than or equal to 1 mm/day) 8
for the periods of 1955-1984 and 1985-2014, respectively. HMD occurrence represented 2.8% and 3.4% of the total number of rain days for the 1955-1984 and 1985-2014 periods, respectively. So while these rain intensities are extremely important in terms of economic and ecological impacts, they do not often explain much of the variability in seasonal rainfall totals. There was very little difference in ENSO phases between the two 30-year periods (Figure 5), so it seems unlikely that the differences in seasonal rainfall variability between the two periods were driven by ENSO. 4. Discussion This study was initiated through stakeholder engagement, and the results have been shared with agricultural stakeholders in the Southeastern U.S. Continuing to learn from and with stakeholders will ensure that meaningful questions are being investigated as we look for solutions for making agriculture and other sectors more resilient to climate risks. The intensity and variability of rainfall is critical for agricultural producers; more variable MAM and JJA rainfall can lead to crop water stress from excessively wet or dry conditions during periods of establishment or during reproductive stages, and high-intensity rains can result in costly losses of soil and nutrients. Fall season (SON) extreme rainfall events might lead to more frequent interruptions in harvest operations or increased soil loss during post-harvest times when there is reduced vegetative cover. Anecdotal reports of heavier rainfall have been heard from producers in the Southeastern U.S. for several years (Bartels et al. 2013), and our study of rainfall intensity during the last 60 years in the Southeastern U.S. showed a significant increase in the numbers of extreme daily rainfall and showed small increases in numbers of very heavy daily rains. While extreme and very heavy rainfall are rare by definition, they are typically high-impact rains, meaning that the results of these rain events in agricultural areas can often be of major economic importance due to the soil and nutrient losses that can accompany extreme rainfall. These findings seem to be consistent with the personal experiences of some producers in the region. Continuous and direct engagement with producers and Extension professionals is important for developing climate risk management research that is relevant to stakeholder interests and management solutions. A novel measure of rainfall intensity, HMD, the occurrence of having greater than or equal to half of the total monthly rainfall in a single day, places rainfall intensity in the context of monthly total rainfall. This quantity was recommended by agricultural producers and is introduced here as a rainfall intensity metric that may be of particular interest to agricultural stakeholders. The HMD measure, demonstrated in this study, allows stakeholders to see how HMD return period varies between months. This is important because a monthly metric can be
9
useful for interpreting rainfall intensity changes during particular times of year that are important for crop development or agricultural operations management. We found that the increases in frequency of extreme daily rainfall were occurring mostly in fall and summer seasons. Tropical cyclones are the meteorological causes of 58% and 71% of the 1-in-5-year rain events in the Southeastern U.S. in summer and fall seasons, respectively (Kunkel et al. 2012). There is no strong observational evidence for trends in the intensity or frequency of tropical cyclones; the substantial fluctuations in frequency and intensity of tropical cyclones make it difficult to robustly detect trends (Knutson et al. 2010; Zwiers et al. 2013). However, there is some evidence of a positive correlation between temperatures and rain intensity (Allan and Soden 2008; Trenberth 2011), which suggests that in a warmer climate, having more column-integrated water vapor (Trenberth et al. 2005; Durre et al. 2009), the rainfall from tropical cyclones could be of greater intensity. This could be part of the explanation for the increased frequency of summer and fall season extreme daily rainfall. Regional average temperature for the region increased at about 0.16⁰C/decade from 1955 to 2014. The average annual temperature was 17.2⁰C for 1955-1984 and 17.8⁰C for 1985-2014. Seasonal increases in temperature between 1955-1984 and 1985-2014 were 1.1, 0.4, 0.6, and 0.5 ⁰C for DJF, MAM, JJA, and SON seasons, respectively. Irrigated area in the region increased more than four-fold from 209,000 ha to 1,120,000 ha between 1954 and 2013 (USDA 2015). There is presently a lot of research aiming to identify the physical mechanisms driving the increased extreme rains in many areas (Kunkel et al 2013), and we have presented temperature and irrigation changes here only as possible contributors. 5. Conclusions An important message from this study is that stakeholders are particularly interested in the timing (what months or seasons) of changing climate risks. Stakeholders also want climate risk research to be closely associated with possible management solutions. Another important message from our findings is that the observed increases in high-intensity rainfall in the Southeastern U.S. could have negative impacts on agriculture if these changes persist. However, there are opportunities for agricultural producers to improve their resilience in a changing world; cover cropping and reduced tillage are among the leading ways that producers in the region are presently adjusting management to improve production and soil health today and to be ready for the possibility of a more challenging rainfall regime. While tillage and cover cropping choices are complex decisions, increasing residue cover of soils might become a greater priority for producers looking to minimize soil and nutrient losses from extreme rains and to capture scarce rainfall in dry seasons. We believe there is a valuable lesson in this experience: exploring climate risk research with stakeholders can provide the opportunity for connecting management solutions with emerging climate challenges. 10
Acknowledgements This research and stakeholder engagement was supported by Cli ate a ia ilit to li ate change: Extension challenges and oppo tu ities i the “outheast U“A, a project funded by Agriculture and Food Research Initiative Competitive Grant no. 2011-67003-30347 from the U ited “tates Depa t e t of Ag i ultu e s National Institute of Food and Agriculture. References Allan RP, Soden BJ (2008) Atmospheric Warming and the Amplification of Precipitation Extremes. Science 321(5895): 1481-1484 Bartels W, Furman CA, Diehl DC, Royce FS, Dourte DR, Ortiz B, Zierden DF, Irani TA, Fraisse CW, Jones JW (2012) Warming up to climate change: A participatory approach to engaging with agricultural stakeholders in the Southeast US. Reg Environ Change. Doi: 10.1007/s10113-0120371-9 Bartels W, Dourte DR, Furman CA (2013) Whe it ai s it pou s E t e e e e ts & ate management in the Tri-state region. Report from 8th Tri-state workshop. http://www.agroclimate.org/seclimate/row-crop-climate-group/ Accessed 22 February 2014. De Angelis AM, Broccoli AJ, Decker SG (2013) A Comparison of CMIP3 Simulations of Precipitation over North America with Observations: Daily Statistics and Circulation Features Accompanying Extreme Events. J Climate 26: 3209-3230 Delgado JA, Groffman PM, Nearing MA, Goddard T, Reicosky D, Lal R, Kitchen NR, Rice CW, Towery D, Salon P (2011) Conservation practices to mitigate and adapt to climate change. J Soil Water Conserv 66(4): 118A-129A Dourte DR, Shukla S, Singh P, Haman D (2012) ‘ai fall I te sit ‐ Du atio ‐F e ue Relationships for Andhra Pradesh, India: Changing Rainfall Patterns and Implications for Runoff and Groundwater Recharge. J Hydrol Eng 18(3): 324-330 Durre I, Williams Jr. CN, Yin X, Vose RS (2009) Radiosonde-based trends in precipitable water over the Northern Hemisphere: An update. J Geophys Res 114: D05112. doi: 10.1029/2008JD010989 Groisman PY, Knight RW, Karl TR (2012) Changes in intense precipitation over the central United States. J Hydrometeorol 13: 47-66 Higgins RW, Kousky VE (2013) Changes in Observed Daily Precipitation over the United States between 1950–79 and 1980–2009. J Hydrometeorol 14: 105-121 Hill H, Hadarits M, Rieger R, Strickert G, Davies EGR, Strobbe KM (2014) The Invitational Drought Tournament: What is it and why is it a useful tool for drought preparedness and adaptation? Weather and Climate Extremes 3: 107-116.
11
Howden SM, Soussana JF, Tubiello FN, Chhetri N, Dunlop M, Meinke H (2007) Adapting agriculture to climate change. Proc Natl Acad Sci USA 104 (50): 19691-19696 Howden SM, Schroeter S, Crimp S, Hanigan I (2014) The changing roles of science in managing Australian droughts: An agricultural perspective, Weather and Climate Extremes 3: 80-89 Karl TR, Melillo JT, Peterson TC (2009) Global Climate Change Impacts in the United States. Karl TR, Melillo JT, Peterson TC, Eds. Cambridge University Press, 189 pp Keggenhoff I, Elizbarashvili M, Amiri-Farahani A, King L (2014) Trends in daily temperature and precipitation extremes over Georgia, 1971–2010. Weather and Climate Extremes 4: 75-85 Kunkel KE, Easterling DR, Kristovich DAR, Gleason B, Stoecker L, Smith R (2012) Meteorological causes of the secular variations in observed extreme precipitation events for the conterminous United States. J Hydrometeor 13: 1131-1141 Kunkel KE et al (2013) Monitoring and understanding trends in extreme storms – state of knowledge. Bull Am Meteorol Soc 94: 499-514 Lal R, Delgado JA, Gulliford J, Nielsen D, Rice CW, Van Pelt RS (2012) Adapting agriculture to drought and extreme events. J Soil Water Conserv 67(6): 153A-157A Langhans C, Govers G, Diels J, Clymans W, Van den Putte A (2010) Dependence of effective hydraulic conductivity on rainfall intensity: loamy agricultural soils. Hydrol Process 24: 22572268 Li W, Li L, Fu R, Deng Y, Wang H (2011) Changes to the North Atlantic Subtropical High and its role in the intensification of summer rainfall variability in the Southeastern United States. J Clim 24: 1499-1506 Lobell DB, Schlenker W, Costa-Roberts J (2011) Climate trends and global crop production since 1980. Science 333(6042): 616-620 Meinke H, Nelson R, Kokic P, Stone R, Selvaraju R, Baethgen W (2006) Climate Res 33: 101-110 Menne MJ, Durre I, Gleason BG, Houston TG, Vose RS (2013) An overview of the Global Historical Climatology Network-Daily database. J Atmos Ocean Tech 29: 897-910 Min S-K, Zhang X, Zwiers FW, Hegerl GC (2011) Human contribution to more-intense precipitation extremes. Nature 470: 378-381 Mirhosseini G, Srivastava P, Stefanova L (2013) The impact of climate change on rainfall Intensity–Duration–Frequency (IDF) curves in Alabama. Reg Environ Change: 13 (Suppl 1): S25S33 Morin J, Benyamini Y (1977) Rainfall infiltration into bare soils. Water Resour Res 13(5): 813817
12
Morris HLC, Megalos MA, Vuola AJ, Adams DC, Monroe MC (2014) Cooperative Extension and climate change: Successful program delivery. J Extension 52(2): Article 2COM3 Knutson TR, McBride JL, Johnny Chan J, Emanuel K, Holland G, Landsea C, Held I, Kossin JP, Srivastava AK, Sugi M (2010) Tropical cyclones and climate change. Nature Geosci 3: 157-163 Sharpley AN (1985) Depth of surface soil-runoff interaction as affected by rainfall, soil slope, and management. Soil Sci Soc Am J 49: 1010-1015 Sillmann J, Kharin VV, Zwiers FW, Zhang X, Bronaugh D (2013) Climate extremes indices in the CMIP5 multimodel ensemble: Part 2. Future climate projections. J Geophys Res-Atmos 118: 2473-2493 Trenberth KE, Dai A, Rasmussen RM, Parsons DB (2003) The changing character of precipitation. Bull Am Meteorol Soc 84: 1205-1217 Trenberth KE, Fasullo J, Smith L (2005) Trends and variability in column integrated atmospheric water vapor. Clim Dynam 24: 741-758 Trenberth KE (2011) Changes in precipitation with climate change. Clim Res 47: 123-138 USDA (2015) United States Department of Agriculture National Agricultural Statistics Service. http://www.agcensus.usda.gov/Publications/Irrigation_Survey/ Accessed 15 January 2015. Walsh J et al (2014) Ch. 2: Our Changing Climate. Climate Change Impacts in the United States: The Third National Climate Assessment, Melillo JM, Terese (T.C.) Richmond TC, Yohe GW, Eds. U.S. Global Change Research Program 19-67. doi: 10.7930/J0KW5CXT Walthall CL et al (2012) Climate Change and Agriculture in the United States: Effects and Adaptation. USDA Technical Bulletin 1935, Washington, DC Wang H, Fu R, Kumar A, Li W (2010) Intensification of summer rainfall variability in the Southeastern United States during recent decades. J Hydrometeorol 11: 1007-1018 Wojcik DJ, Monroe C, Adams DC, Plate RR (2014) Message in a bottleneck? Attitudes and perceptions of climate change in the Cooperative Extension Service in the southeastern United States. J Hum Sci Extension 2(1): 51-70 Zwiers FW, Alexander LV, Hegerl GC, Knutson TR, Kossin JP, Naveau P, Nicholls N, Schär C, Seneviratne SI, Zhang X (2013) Climate extremes: Challenges in estimating and understanding recent changes in the frequency and intensity of extreme climate and weather events. In: Asrar GR, Hurrell JW (eds) Climate science for serving society: Research, modelling and prediction priorities. Springer, Geneva, pp 339–390
13
Table captions Table I Daily rainfall amounts for five descriptive categories of rain events based on responses from groups of Producers, Researchers, and Extension. Right column indicates the thresholds used in this study.
14
Figure captions Fig. 1 Station-specific changes between 1985-2014 and 1955-1984 in occurrence of extreme daily rain ( > 152.4 mm/day) for the 259 stations across that Southeastern U.S. Light-color dots correspond to decreases and dark-color dots correspond to increases in extreme rain events. The maximum change, mean of all differences, and minimum change are 7, 0.8, and -4.
Fig. 2 Southeastern U.S. average percent change in numbers of rain events in the four fixed-threshold rainfall categories: percent change shows difference in numbers of fixed-threshold rains in the most recent 30-year period (1985-2014) compared to the previous 30-year period (1955-1984). Solid fill in the figure shows statistically significant differences between the two time periods.
Fig. 3 Southeastern U.S. seasonally-averaged percent change in numbers of fixed-threshold rain in the four fixed-threshold rainfall categories for four different seasons (winter, DJF; spring, MAM; summer, JJA; fall SON). Percent change shows difference between the periods of 1985-2014 and 1955-1984. Solid fill in the figure shows statistically significant differences between the two time periods.
Fig. 4 Southeastern U.S. station-average return periods and percent change in return periods (between the 1955-1984 and 1985-2014 periods) of having at least half of monthly rain total in single day. Significant differences, designated by solid fill in the figures, were found in numbers of months in which at least half of monthly total rain occurred in a single day (HMD) for all months except June, August, September, and December.
Fig. 5 Seasonal rainfall anomalies in DJF, MAM, JJA, SON for the Southeastern U.S. Figures show the number of years in which total seasonal rainfall exceeded 1 standard deviation from the average for the 30-year periods of 1955-1984 and 1985-2014. Spring and summer seasons show increased variability in the last 30 years compared to the previous 30 years, while winter and fall show decreased rainfall variability in the last 30 years compared to the previous 30 years. ENSO phases are similar between the two 30-year periods
15
16
17
18
19
20
Daily rain description Good Moderately heavy Heavy Very Heavy Extreme
Daily rain amounts (mm) Producers Researcher 12.7 (early), 38.1 (reproductive) 12.7 to 25.4 50.8 25.4 to 50.8 76.2 50.8 to 76.2 101.6 76.2 to 101.3 > 152.4 > 101.6
Extension 12.7 to 25.4 38.1 76.2 127 > 152.4
Highlights
High-intensity rains can be costly in agricultural systems due to soil/nutrient loss There have been significantly more extreme rain events in the Southeastern U.S. A new monthly rain intensity metric is introduced based on stakeholder input Stakeholder engagement can help improve the links between research and practice
21
This Study 12.7 to 25.4 25.4 to 76.2 ≥ . ≥ .
