Published July 14, 2014

Journal of Environmental Quality

Special Section

Long-Term Agro-Hydrological Research in UWR Experimental Watersheds

Climate Change and Observed Climate Trends in the Fort Cobb Experimental Watershed J. D. Garbrecht,* X. C. Zhang, and J. L. Steiner

W

eather, climate variability, and climate change are important considerations in water resources related investigations, such as urban water supply projections, environmental quality assessments (Delgado et al., 2011), agricultural production planning (Hatfield et al., 2011; Jones et al., 2000), land resources management (Nearing et al., 2004), and development of conservation and preparedness strategies (Climate Change Science Program, 2008; Lal et al., 2011). In this context, drought is of particular relevance because it often results in substantial, far-reaching, and lasting impacts on the environment and water-dependent sectors of the economy. Drought is characterized by a gradual accumulation of deficits in precipitation and water supply at local to regional scales (National Climate Assessment and Development Advisory Committee, 2009; American Meteorological Society, 2013). Drought is a naturally occurring, temporary feature of the climate cycle (Hoerling et al., 2013a). In this study, year and multiyear temporary and naturally occurring climate departures from normal are referred to as climate variations (Garbrecht, 2008, 2011). The occurrence and severity of drought have been linked to natural climate system variability, including sea surface temperature oscillations and polar jet stream meandering (Peterson et al., 2013; Hoerling et al., 2013b). It is estimated that drought accounts for roughly a quarter of all losses from major weather events (American Meteorological Society, 2013). Agriculture is generally the first sector to be impacted and is especially susceptible to short-term droughts of 6 mo or less (Illston and Basara, 2003). It has also been recognized that the climate system is slowly warming. Much of the climate warming during the past 50 yr has been attributed to human activity, primarily in the form of increasing heat-trapping greenhouse gases in the atmosphere (National Climate Assessment and Development Advisory Committee, 2009). The impacts of a warming climate are more permanent because they will persist for decades due to greenhouse gases already added to the atmosphere. Drought frequency, duration, and severity resulting from natural climate variability are likely to increase in a warmer climate as evaporation increases and regional and seasonal precipitation and temperature patterns change (American Meteorological Society, 2013; Karl et al., 2009; Hoerling et al., 2013b). In this

Recurring droughts in the Southern Great Plains of the United States are stressing the landscape, increasing uncertainty and risk in agricultural production, and impeding optimal agronomic management of crop, pasture, and grazing systems. The distinct possibility that the severity of recent droughts may be related to a greenhouse-gas induced climate change introduces new challenges for water resources managers because the intensification of droughts could represent a permanent feature of the future climate. Climate records of the Fort Cobb watershed in central Oklahoma were analyzed to determine if recent decade-long trends in precipitation and air temperature were consistent with climate change projections for central Oklahoma. The historical precipitation record did not reveal any compelling evidence that the recent 20-yr-long decline in precipitation was related to climate change. Also, precipitation projections by global circulation models (GCMs) displayed a flat pattern through the end of the 21st century. Neither observed nor projected precipitation displayed a multidecadal monotonic rising or declining trend consistent with an ongoing warming climate. The recent trend in observed annual precipitation was probably a decade-scale variation not directly related to the warming climate. On the other hand, the observed monotonic warming trend of 0.34°C decade−1 that started around 1978 is consistent with GCM projections of increasing temperature for central Oklahoma.

Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

USDA–ARS, Grazinglands Research Lab., 7207 West Cheyenne St., El Reno, OK 73036. Assigned to Associate Editor Claire Baffaut.

J. Environ. Qual. 43:1319–1327 (2014) doi:10.2134/jeq2013.07.0286 Supplemental data files are available online for this article. Received 22 July 2013. *Corresponding author ([email protected]).

Abbreviations: CMIP3, Coupled Model Intercomparison Project Phase 3; GCM, global circulation model; GHG, greenhouse gas; WCRP, World Climate Research Programme; WMA, weighted moving average.

1319

study, multidecadal to century-long departures from normal due to the accumulation of greenhouse gases in the atmosphere are referred to as climate trend or climate change. Climate models, called global circulation models (GCMs), are used to project potential climate change for assumed future greenhouse gas emission scenarios. Projected climate trends are at regional to continental and decade to century scales. Global circulation model simulations suggest that during the next decades drought will increase across much of North America, climate-system trends observed in recent decades will continue, and a tendency toward more intense rain events separated by longer periods without precipitation will persist (American Meteorological Society, 2012; Karl et al., 2009; National Climate Assessment and Development Advisory Committee, 2009). It is important to recognize that climate trends at the continental scale do not necessarily translate into similar climate trends at the regional scale, and climate trends are not all uniform across the United States. The climate analysis and findings presented here are specific to the central Oklahoma region. In the Southern Great Plains, the climate is highly variable. Seasonal, year to year, and persistent multiyear variations in precipitation and air temperature are the norm (Garbrecht and Rossel, 2002; Garbrecht et al., 2004). Agricultural droughts occur frequently, and their impacts may linger long after the initial causes leading to drought have ended (National Climate Assessment and Development Advisory Committee, 2009; American Meteorological Society, 2013). The possibility that the recent occurrence and severity of droughts in the Southern Great Plains (Hoerling et al., 2013b) may be related, at least in part, to a changing climate introduces another dimension and challenge to long-term water and land resource management, where decisions are primarily made at local to watershed and year to decade scales (Karl et al., 2009). From this application perspective, the relevant questions are whether changes in climate can be identified in the observed climate record at the watershed and decade scales, which components of the climate are changing, at what rate, and what the climate may look like a decade or two from now. In this study, seasonal and annual precipitation and air temperature records of the Fort Cobb watershed in westcentral Oklahoma were analyzed to shed light on these practical questions often asked by agricultural producers and land managers. Our objectives were to identify recent decade-long trends in the observed precipitation and air temperature of the Fort Cobb watershed and to determine if these trends can be interpreted to be an expression of climate change. Information on the nature and potential persistence of the recent droughts, as well as on climatic characteristics that may prevail a decade from now, will enable agricultural producers and natural resource managers in the region to develop adaptive and mitigating measures. Such measures would increase the resilience of the agricultural landscape to climate change, conserve the natural resource base, preserve environmental quality, and continue the delivery of goods and services from the land under future climatic conditions. Establishing the causality of drought in terms of global circulation mechanisms, land–ocean– atmosphere system behavior, and analyses of weather variables other than precipitation and air temperature (such as dew point, temperature, pan evaporation, wind, and tropospheric pressure) are outside the scope of this study. 1320

Materials and Methods Physiographic characteristics of the watershed, time series of historical climate observations, treatment of gaps and missing values in the climate record, potential future climate outcomes, data sources, and pertinent analysis methodologies were used in this study.

The Fort Cobb Reservoir Watershed The Fort Cobb Reservoir in west-central Oklahoma was constructed in 1958 to 1959 by the U.S. Bureau of Reclamation and serves the multiple purposes of flood control, municipal and industrial water supply, wildlife habitat, and recreation. The drainage area above the dam is 787 km2. The reservoir flood pool and flood releases are regulated by the U.S. Army Corps of Engineers, and the conservation pool and water supply withdrawals are managed by the Fort Cobb Reservoir Master Conservancy District. Land use in the watershed consists primarily of crop and range or pasture land. Based on the 1983 to 2012 time period, mean annual precipitation is about 780 mm yr−1. May and June are the two wettest months of the year and account for about 30% of the annual precipitation. The coldest month of the year is January and the warmest is July, with monthly average air temperatures of 3 and 28°C, respectively. The watershed was selected as a research site because the reservoir was listed by the Oklahoma Department of Environmental Quality (ODEQ) as not meeting the state’s water quality standards based on hypereutrophication of the reservoir associated with nutrient loading from the agricultural watershed. The ODEQ finalized a total maximum daily load plan to mitigate water quality concerns in 2007 based on P loading. While nutrient loading to the reservoir was not the focus of this study, the climate issues addressed in this study are highly relevant for water quality investigations. Climate, weather, and floods and droughts affect the hydrologic cycle and indirectly runoff, which in turn determines the movement and fate of sediment and nutrients throughout the watershed.

Climate Data The climate trend analysis was based on monthly and annual precipitation and air temperature records of the Fort Cobb watershed and Fort Cobb climate division and on climate change projections. The climate record of the Fort Cobb watershed was calculated from daily observations at weather stations located in the vicinity of the watershed. The climate record for the Fort Cobb climate division was based on interpolated data of neighboring NOAA climate divisions (Guttman and Quayle, 1996; National Climate Data Center, 2012). Climate divisions involved in this interpolation are shown in Fig. 1 and included the West Central (OK 3404), Central (OK 3405), and South West (OK 3407) climate divisions. The climate change projections for the Fort Cobb region were extracted from the archive of the GCM projections of the World Climate Research Programme’s (WCRP) Coupled Model Intercomparison Project Phase 3 (CMIP3) multimodel data set (Maurer et al., 2007).

Climate Record of the Fort Cobb Watershed A continuous 64-yr long daily weather record (1949–2012) was developed at five weather stations around the Fort Cobb Journal of Environmental Quality

Fig. 1. Climate Divisions of Oklahoma, situation map of the Fort Cobb Reservoir watershed, and location of relevant National Weather Service Cooperative Observer Network (COOP), Mesonet (mesoscale weather observation network), and Micronet (microscale weather observation network) weather stations.

ratio of the means of the weather variable at the primary and support stations. The calculation of the adjustment factor and a data quality control for stations belonging to the USHCN are detailed in Supplement 1, Text S3. The meteorological database of the VIC model started in 1949 and could not be drawn on to estimate missing values before 1949. Therefore, the starting date for the weather data at the five primary stations was set to 1949. The continuous, gap-free, and quality controlled 1949– 2012 daily weather record at the five primary stations is given in Supplement 2 (Primary_Stations.xlsx). The daily precipitation at the Weatherford, Carnegie, Fort Cobb and Lookeba COOP stations were first averaged in space using the Thiessen method (Thiessen, 1911; cited in Chow, 1964) to produce daily average precipitation over the Fort Cobb watershed. The latter were summed to monthly and annual precipitation at the watershed scale. Likewise, daily average air temperature at the Weatherford and Carnegie COOP and Hinton Mesonet stations were first averaged in space using station weights determined by the Thiessen method. The resulting daily temperature was then averaged into monthly and annual temperature at the watershed scale. These monthly and annual precipitation and air temperature records for the Fort Cobb watershed were the basis to identify recent climate trends

Reservoir watershed. The five stations, called primary stations, were those at the towns Weatherford, Carnegie, Lookeba, Hinton, and Fort Cobb (Fig. 1; Table 1). Additional details regarding primary stations, such as location, completeness of record, and estimation of missing weather observations are given in Supplement 1, Text S1. Gaps and sporadic missing values in the daily precipitation and air temperature records of the primary stations were filled with adjusted values from other nearby stations, called support stations. Support stations include National Weather Service Cooperative Observer Network (COOP) stations (Fiebrich and Crawford, 2009), Oklahoma Mesonet stations (McPherson et al., 2007), ARS Micronet stations (Elliott et al., 1993), U.S. Historical Climate Network (USHCN) stations (Menne et al., 2009), and the 1949–2010 1/8th degree meteorological forcing data of the variable infiltration capacity (VIC; Liang et al., 1994) model database. Additional details regarding support stations, acronyms, references, and sources of weather data are given in Supplement 1, Text S2. Transfer of weather data from a support to a primary station requires adjustments to correct for spatial weather gradients, changes in station altitude and exposure, or any other causes that may introduce a systematic bias. The adjustment factor was the

Table 1. Primary meteorological station information from is National Weather Service Cooperative Observer Network (COOP) and the Mesonet stations. Primary station name Weatherford

Station type COOP

Station ID 34049422

Carnegie Fort Cobb

COOP COOP

34071504 34073281

Observed variable† P T P&T P

Lookeba Hinton

COOP Mesonet

34075329 HINT

P T

Available weather data 1905–2012 1905–7 Feb. 2011 1914–30 Sept. 2005 1938–31 Aug. 1975 1940–31 Oct. 2010 1 Mar. 1997–2012

Data extension‡ Mesonet WEAT Micronet F115 Anadarko COOP Mesonet FTCB Micronet F111 VIC

† P, precipitation; T, air temperature. ‡ Name of meteorological station used to extend weather record past closure date of primary station. The Mesonet is a mesoscale weather observation network (McPherson et al., 2007); the Micronet is a microscale weather observation network (Elliott et al., 1993; Starks et al., 2014); VIC is the variable infiltration capacity model (Liang et al., 1994). www.agronomy.org • www.crops.org • www.soils.org 1321

in the Fort Cobb watershed region. Climate characteristics of the Fort Cobb watershed over the most recent 30 yr (1983–2012) were assumed to be representative of current climate conditions and were chosen to be the common denominator to which all climate projections were referenced. The monthly and annual precipitation and air temperature of the Fort Cobb watershed are given in Supplement 3 (Fort_Cobb_Climate.dat).

Climate Record for the Fort Cobb Climate Division In addition to the gap-free 1949 to 2012 daily weather records at the five primary stations and the calculated monthly and annual precipitation and temperature records of the Fort Cobb watershed, a 1895 to 2012 monthly climate record for the Fort Cobb watershed region was developed using data from the U.S. Climate Divisional Database (Guttman and Quayle, 1996) of the National Climatic Data Center (2012). Monthly precipitation and air temperature from three Oklahoma climate divisions (OK3404, OK3405, and OK3407) that surround the watershed were interpolated onto the watershed. Interpolated monthly and annual climate values were adjusted up or down so that the 1983 to 2012 mean matched the reference mean of the Fort Cobb watershed. The resulting climate record is referred to as that of the Fort Cobb division. The climate record of the Fort Cobb division is twice as long as the 1949 to 2012 daily weather record and provides a snapshot of climate trends and variations back to 1895. It was used in time-series plots of climate variables to visualize and assess the magnitude of recent climate trends in the Fort Cobb watershed in relation to variations in the longer historical context. In this study, interpolated climate division data are referred to as Fort Cobb division data and the spatial scale as the division scale.

Climate Change Projections General circulation models have been used to projected monthly precipitation and air temperature out to the end of the 21st century. The archive of the GCM projections is the WCRP’s CMIP3 multimodel data set. Climate projections were spatially disaggregated and bias corrected following the technique described by Maurer et al. (2007). Climate projections of seven GCMs for each of three future greenhouse gas (GHG) emission scenarios were retrieved from the CMIP3 archive for the Fort Cobb watershed. The seven projections for each scenario were assumed to be equally likely future climate outcomes and their average was assumed to be representative of the actual, but unknown, climate through the year 2100. The three GHG emission scenarios that were selected for this study are the higher, middle, and lower emission paths, as defined by the Intergovernmental Panel on Climate Change (2000). Specifics on the GCMs and GHG emission scenarios are given in Supplement 1, Tables S1 and S2.

Trend Line A climate trend is a slow, monotonic change in climate across a large region and long duration that displays low-frequency climate variations representing the nonstationary nature of the climate. 1322

A climate trend line was defined as the 31-yr weighted moving average (WMA) of a time series of annual or seasonal climate values. The weights of the 31-yr WMA are highest at the center of the moving average window and decrease away from the center toward the two ends of the moving average window. The WMA filtered out most of the transient climate variations (year- to decade-scale departures from the mean) and resulted in a smooth trend-line curve that followed persistent and systematic changes in climate (decade- to century-scale departures from the mean). In this study, an actual climate trend was assumed to be present when the trend line exhibited a persistent monotonic increase or decline lasting 30 yr or longer. The rate of change of a trend was approximated as the slope of a linear regression fitted to the 5-yr WMA of the climate variable over the duration of the trend. A 5-yr WMA was used to dampen the effects of occasional high-influence and high-leverage outlier data points. Such outlier data points are not unusual in precipitation records.

Results

Annual Climate Trend Analysis Precipitation

The observed watershed- and division-scale annual precipitation were highly correlated (r2 = 0.88; based on 1949– 2012 time period), suggesting that annual precipitation at the division scale is a good indicator of what the watershed-scale precipitation would have been if it had been observed before 1949. A plot of the annual precipitation and corresponding trend lines shows that from 1895 to 1970 the precipitation trend line was essentially flat at about 690 mm (Fig. 2, left side, blue line), rose in the 1970s and 1980s to about 810 mm, and decreased in the two subsequent decades to the pre-1970 level (Fig. 2, black

Fig. 2. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected annual precipitation. Blue lines are 1895 to 2012 Fort Cobb Division precipitation. Black lines are 1949 to 2012 Fort Cobb watershed precipitation. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 23-yr rise in precipitation starting in 1973 and a 15-yr decline in precipitation starting in 1997. The thin horizontal black line is the 1983 to 2012 reference watershed precipitation. The thick red, gold, and green lines are trend lines of the averaged CMIP3 precipitation projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions), respectively. The colored thin lines define the 10th and 90th percentiles of the individual global circulation models precipitation projections for each scenario. Journal of Environmental Quality

and blue lines; also Supplement 1, Fig. S1). Furthermore, a plot of the CMIP3 projected annual precipitation (average of seven GCM simulations) for each of the three emission scenarios also revealed a flat precipitation trend line through the year 2100 (Fig. 2, right side, thick lines). The colored thin lines on the

right-hand side approximate the 80th percentile range of the trend lines of individual GCM annual precipitation projections. Given this context and assuming that the CMIP3 projected annual precipitation is accurate, the upward and then declining trend in annual precipitation during the last four decades can only be interpreted as a temporary departure from long-term pre-1970 average precipitation.

Temperature

Fig. 3. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected annual average air temperature. Blue lines are 1895 to 2012 Fort Cobb Division temperature. Black lines are 1949 to 2012 Fort Cobb watershed temperature. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 35-yr rising trend in observed air temperature starting in 1978. The thin horizontal black line is the 1983 to 2012 reference watershed temperature. The thick red, gold, and green lines are CMIP3 temperature projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The thin red lines define the 10th and 90th percentiles of the individual global circulation models precipitation projections for Emission Scenario A2. For clarity, the percentile lines for Emission Scenarios A1b and B1 have been omitted, but the 80th percentile range is comparable in size to that of Emission Scenario A2.

Annual average air temperatures at the watershed and division scales were also highly correlated (r2 = 0.96; based on 1949–2012 time period), making annual temperature at the division scale a good indicator of annual temperature at the watershed scale and back in time to 1895. A plot of the annual temperature shows an undulating trend-line pattern at the decadal time scale (Fig. 3, left side; also Supplement 1, Fig. S2): an increasing temperature trend of about 0.36°C decade−1 starting in the early 1900s through the early 1930s, followed by a decreasing trend of about 0.15°C decade−1 through the early 1970s, and again an increasing trend after 1977 of about 0.34°C decade−1. This most recent 35-yr (1978–2012) temperature trend (Fig. 3, black line; Supplement 1, Fig. S2) is different from previous trends: (i) the trend line is, at its peak, the highest on record (16.1°C in 2012), surpassing that during the Dust Bowl (15.6°C in 1935); (ii) the trend period includes three of the four highest annual temperatures on record; (iii) the average temperature of the most recent 10 and 20 yr (16.0 and 15.8°C, respectively) are the highest on record; and (iv) after 1998 the annual temperature record contains only values above 15.0°C. The observed annual temperature trend of the Fort Cobb watershed (about 0.34°C decade−1 for 1978–2012) closely followed the trend of the CMIP3 projections of annual temperature for the three GHG emission scenarios (about 0.35°C decade−1 for 1978–2012) (Fig. 3). Based on these findings, it can be stated with some degree of confidence that the 1978 to 2012 increasing observed annual temperature trend of the Fort Cobb watershed is consistent with the CMIP3 projected temperature and is probably a reflection of climate change. Assuming that this is true, the temperature will continue to increase in the near future following the CMIP3 projected temperature trajectories.

Monthly and Seasonal Trend Analysis Precipitation Fig. 4. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected precipitation for March. Blue lines represent 1895 to 2012 Fort Cobb Division precipitation. Black lines represent 1949 to 2012 Fort Cobb watershed precipitation. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 50-yr rising trend in observed precipitation starting in 1957. The thin horizontal black line is the 1983 to 2012 reference watershed precipitation. The thick red, gold, and green lines are CMIP3 precipitation projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The colored thin lines define the 10th and 90th percentiles of the individual global circulation models (GCMs) precipitation projections for each scenario.

Before about 1950, the trend lines of observed monthly precipitation were relatively flat for most calendar months (Supplement 1, Fig. S3). From about 1950 to 2010, the months of March and August showed a statistically significant, 60-yr-long increasing trend in observed precipitation (rate of change of 7.1 and 8.3 mm decade−1, respectively; thick black line in Fig. 4 and 5). However, the trend

www.agronomy.org • www.crops.org • www.soils.org 1323

Fig. S6). These 6 mo were appended into one 6-mo-long time period (March–August), called the spring–summer season. Spring–summer temperatures displayed an unusual increasing temperature trend that started around 1970 and persisted through 2012, as illustrated by the thick black line in Fig. 8 (also Supplement 1, Fig. S7). The 1970 to 2012 increasing trend was statistically significant (at a = 0.01) and had an average rate of increase of 0.34°C decade−1 during the 43-yr duration of the trend. If the trend was considered for only the last 30 yr (1983–2012), the trend would be statistically significant at 0.01 and have an average rate of increase of 0.57°C decade−1. Also noteworthy is that three of the four highest spring–summer temperatures on record occurred in the last 10 yr. The CMIP3 projected temperature for the Fig. 5. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) prospring–summer season for the three GHG emission jected precipitation for August. Blue lines represent 1895 to 2012 Fort Cobb Division precipitation. Black lines represent 1949 to 2012 Fort Cobb watershed precipitation. scenarios (A1b, A2, and B1) are shown on the rightThick lines are trend lines (31-yr weighted moving average). The thick black trend line hand side of Fig. 8 (thick red, yellow, and green shows a 43-yr rising trend in observed precipitation starting in 1964. The thin horitrend lines). The trend lines for both the observed zontal black line is the 1983 to 2012 reference watershed precipitation. The thick red, gold, and green lines are CMIP3 precipitation projections for Emission Scenarios A2 and projected temperatures agreed closely during (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG most of the 1949 to 2012 time period (there was emissions). The colored thin lines define the 10th and 90th percentiles of the individual no relevant difference in the trend lines among global circulation models precipitation projections for each scenario. the three emission scenarios through 2030). The lines of CMIP3 precipitation projections for March and August trend of the observed spring–summer temperatures did not show signs of a rising trend for any of the three emission was 0.34°C decade−1 during the 1970 to 2012 time period and scenarios (Fig. 4 and 5). Thus, based on precipitation projections, 0.28°C decade−1 during the same time period for the CMIP3 there is not enough evidence to infer that the trend in the projections. observed March and August precipitation is or is not related to the warming of the climate. During the last 20 yr (1993–2012), several calendar months displayed a declining precipitation trend (Supplement 1, Fig. S3). First, precipitation in April to June, the historically wettest time period of the year, declined at a rate of 71.42 mm decade−1 (Fig. 6, black line; also Supplement 1, Fig. S4). Second, precipitation in November to February, the driest time period of the year, declined at a rate of 33.0 mm decade−1 (Fig. 7, black line; also Supplement 1, Fig. S5). However, the trend lines of corresponding CMIP3 precipitation projections did not show signs of a declining precipitation trend (Fig. 6 and 7, righthand side). Given the short duration (20 yr) of the declining trend in observed precipitation and the weak statistical significance of the trend, it remains inconclusive if the observed trend is linked to climate change or just part of a decadal precipitation variation.

Air Temperature An increasing trend in the monthly average air temperature was found for March through August. For March and July, the trend started in the 1960s, for May and August in the 1970s, and for April and June in the 1980s. All six monthly trends persisted through 2012 (Supplement 1, 1324

Fig. 6. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected precipitation for April, May, and June. Blue lines represent 1895 to 2012 Fort Cobb Division precipitation. Black lines represent 1949 to 2012 Fort Cobb watershed precipitation. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 16-yr decline starting in 1950, a 16-yr rise starting in 1973, and a 16-yr decline starting in 1996. The thin horizontal black line is the 1983 to 2012 reference watershed precipitation. The thick red, gold, and green lines are CMIP3 precipitation projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The colored thin lines define the 10th and 90th percentiles of the individual global circulation models precipitation projections for each scenario.

Journal of Environmental Quality

The observed air temperature in the fall– winter season (September–February) displayed an increasing trend starting in 1980, but the trend flattened out starting in 2000 (Fig. 9; also Supplement 1, Fig. S8). In contrast, the corresponding trend of the CMIP3 projections persisted through 2012 and beyond (Fig. 9). The agreement in the spring–summer temperature trend and to a lesser degree the fall–winter trend between observed and CMIP3 projected data is consistent with the previous interpretation that sustained increases in the annual temperature of the Fort Cobb watershed and Fort Cobb division during the 1970 to 2012 period (Fig. 3) are probably an expression of GHG-induced climate change. Furthermore, assuming that the increase in annual temperature is in fact due to climate change, then the seasonal air temperature trend analysis (Fig. 8 and 9) suggests that the warming in the last decade has occurred predominantly in the spring–summer season and to a lesser degree in the fall–winter season.

Discussion and Conclusions

Fig. 7. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected precipitation for November, December, January, and February. Blue lines represent 1895 to 2012 Fort Cobb Division precipitation. Black lines represent 1949 to 2012 Fort Cobb watershed precipitation. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 34-yr rising trend in observed precipitation starting in 1959 and a 17-yr decline starting in 1994. The thin horizontal black line is the 1983 to 2012 reference watershed precipitation. The thick red, gold, and green lines are CMIP3 precipitation projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The colored thin lines define the 10th and 90th percentiles of the individual global circulation models precipitation projections for each scenario.

The 1949 to 2012 precipitation and air temperature records of the Fort Cobb watershed and the Fort Cobb climate division were analyzed to identify persistent precipitation and air temperature trends in recent decades and infer if these trends could be attributed, at least in part, to GHG-induced climate change. A 31-yr WMA was applied to the time series of observed precipitation and air temperature to filter out year- to decadescale transient climate variations of mostly natural origin and uncover potential decade- to centuryscale climate trends, mostly due to GHG-induced climate change. Downscaled CMIP3 modeled climate projections were also smoothed, and the resulting trend line was considered to represent precipitation and air temperature trends due to GHG emissions. Consistent trends between observed and simulated precipitation and air temperature would be an indication that the trend in observed data was probably an expression of climate change and that the trend would persist for decades to come. Fig. 8. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) proPrecipitation records of the Fort Cobb jected average air temperatures for the spring and summer seasons (March–August). watershed exhibited an increasing trend line Blue lines represent 1895 to 2012 Fort Cobb Division temperature. Black lines represent in annual precipitation in the 1970s and 1949 to 2012 Fort Cobb watershed temperature. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 38-yr rising trend in 1980s, followed by a declining trend line in the observed air temperature starting in 1976. The thin horizontal black line is the 1983 subsequent two decades, resulting in current to 2012 reference watershed temperature. The thick red, gold, and green lines are annual precipitation characteristics similar to CMIP3 temperature projections for Emission Scenarios A2 (high greenhouse gas [GHG] emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The thin red those before 1970. This pattern of the trend line lines define the 10th and 90th percentiles of the individual global circulation models suggests that the precipitation departure from precipitation projections for Emission Scenario A2. For clarity, the percentile lines for normal is a temporary variation and probably does Emission Scenarios A1b and B1 have been omitted, but the 80th percentile range is comparable in size to that of Emission Scenario A2. not represent an enduring and lasting trend that is related to climate change. This interpretation projections, which were trend free through 2100. At the was supported by the CMIP3 modeled annual precipitation seasonal time scale (3 consecutive mo), relevant seasonal www.agronomy.org • www.crops.org • www.soils.org 1325

precipitation trends could not be found in the Fort Cobb watershed precipitation record nor were any found in the CMIP3 seasonal precipitation projections. At the monthly time scale, March and August precipitation displayed a statistically significant 58-yr-long (1950–2007) rising trend of 7 to 8 mm decade−1. The trend appears to flatten out after year 2007 (Fig. 4 and 5). A corresponding trend in the CMIP3 projected precipitation was not found. Also, the observed March and August precipitation trend does not appear to be linked to long-term changes in the dominant atmospheric mechanisms that control the climate in Oklahoma (polar jet stream [Francis and Vavrus, 2012] and North Atlantic subtropical high [Li et al., 2011, 2013]). Thus, no conclusive evidence was found in the historical precipitation record that permitted us to infer whether or not recent monthly, seasonal, and Fig. 9. Observed and Coupled Model Intercomparison Project Phase 3 (CMIP3) projected average air temperatures for the fall and winter seasons (September–February). annual precipitation trends were enduring features Blue lines represent 1895 to 2012 Fort Cobb Division temperature. Black lines represent related to, associated with, or representative of 1949 to 2012 Fort Cobb watershed temperature. Thick lines are trend lines (31-yr weighted moving average). The thick black trend line shows a 23-yr decline starting climate change. in 1956 and a 25-yr decline starting in 1981. The thin horizontal black line is the 1983 The observed annual average air temperature to 2012 reference watershed temperature. The thick red, gold, and green lines are of the Fort Cobb watershed displayed a persistent CMIP3 temperature projections for Emission Scenario A2 (high greenhouse gas [GHG] 40-yr-long warming trend (0.34°C decade−1) emissions), A1b (middle GHG emissions), and B1 (low GHG emissions). The thin red lines define the 10th and 90th percentiles of the individual global circulation models that started in 1972 and has persisted through precipitation projections for Emission Scenario A2. For clarity, the percentile lines for the present time. This warming trend closely Emission Scenarios A1b and B1 have been omitted, but the 80th percentile range is followed the warming trend of the CMIP3 comparable in size to that of Emission Scenario A2. model projected temperature for the three References GHG emission scenarios (0.35°C decade−1). At a seasonal American Meteorological Society. 2012. Climate change: An information time scale, the spring–summer season exhibited a 1970 to statement of the American Meteorological Society. AMS, Boston. http:// 2012 warming trend, which was close to the temperature ametsoc.org/POLICY/2012climatechange.html (accessed 3 Feb. 2014). trend projected by the CMIP3 models. The close agreement American Meteorological Society. 2013. Drought: An information statement of the American Meteorological Society. AMS, Boston. http://ametsoc.org/ between observed and projected annual and spring–summerPOLICY/2013drought_amsstatement.html (accessed 3 Feb. 2014). season temperatures suggests that recent temperature trends Anandhi, A., S. Perumal, P.H. Gowda, M. Knapp, S. Hutchinson, J. may be a sign of climate change. Furthermore, the flat trend Harrington, Jr., et al. 2013. Long-term spatial and temporal trends in frost indices in Kansas, USA. Clim. Change 120:169–181. doi:10.1007/ line of the fall–winter season that started in 2000 suggests that s10584-013-0794-4 an above-average portion of the annual warming trend occurs Chow, V.T. 1964. Handbook of applied hydrology. McGraw-Hill Book Co., in the spring–summer season. The warming trend in the Fort New York. Cobb watershed region is in agreement with the findings of a Climate Change Science Program. 2008. The effects of climate change on agriculture, land resources, water resources and biodiversity. Synthesis and recent investigation of frost indices in Kansas that concluded Assessment Product 4.3. USEPA, Washington, DC. that the state was experiencing a long and gradual warming Delgado, J.A., P. Groffman, M.A. Nearing, T. Goddard, D. Reicosky, R. Lal, et trend (Anandhi et al., 2013). al. 2011. Conservation practices to mitigate and adapt to climate change. J. Soil Water Conserv. 66:118A–129A. doi:10.2489/jswc.66.4.118A In summary, agricultural producers and water resources Elliott, R.L., F.R. Schiebe, K.C. Crawford, K.D. Peter, and W.E. Puckett. 1993. managers in central Oklahoma will probably see little change A unique data capability for natural resources studies. Paper presented at in annual and seasonal rainfall amounts due to climate change. the International Winter Meeting of the American Society of Agricultural However, the observed recent warming trend of 0.34°C decade−1 Engineering, Chicago. 14–17 Dec. 1993. Paper 932529. Fiebrich, C.A., and K.C. Crawford. 2009. Automation: A step toward is consistent with GCM-modeled climate projections, which improving the quality of daily temperature data produced by climate indicate a continued warming in the near future. observing networks. J. Atmos. Ocean. Technol. 26:1246–1260.

Acknowledgments We acknowledge the modeling groups, the Program for Climate Model Diagnosis and Intercomparison (PCMDI) and the WCRP’s Working Group on Coupled Modelling (WGCM), for their roles in making available the WCRP’s CMIP3 multimodel data set. Support for this CMIP3 data set was provided by the Office of Science, U.S. Department of Energy.

1326

doi:10.1175/2009JTECHA1241.1 Francis, J.A., and S.J. Vavrus. 2012. Evidence linking artic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett. 39(6):L06801. doi:10.1029/2012GL051000 Garbrecht, J.D. 2008. Effects of multiyear precipitation variations on watershed runoff and sediment yield. J. Soil Water Conserv. 63:52A–53A. doi:10.2489/jswc.63.2.52A Garbrecht, J.D. 2011. Effects of climate variations and soil conservation on sedimentation of a west-central Oklahoma reservoir. J. Hydrol. Eng. 16:899–913. doi:10.1061/(ASCE)HE.1943-5584.0000377

Journal of Environmental Quality

Garbrecht, J.D., and F. Rossel. 2002. Decade-scale precipitation increase in the Great Plains at the End of the 20th Century. J. Hydrol. Eng. 7:64–75. doi:10.1061/(ASCE)1084-0699(2002)7:1(64) Garbrecht, J., M. Van Liew, and G.O. Brown. 2004. Trends in precipitation, streamflow and evapotranspiration in the Great Plains of the United States. J. Hydrol. Eng. 9:360–367. doi:10.1061/(ASCE)1084-0699(2004)9:5(360) Guttman, N.V., and R.G. Quayle. 1996. A historical perspective of U.S. climate divisions. Bull. Am. Meteorol. Soc. 77:293–303. doi:10.1175/1520-0477(1996)0772.0.CO;2 Hatfield, J.L., K.J. Boote, B.A. Kimball, L. Ziska, R.C. Izaurralde, D. Ort, et al.. 2011. Climate impacts on agriculture: Implications for crop production. Agron. J. 103:351–370. doi:10.2134/agronj2010.0303 Hoerling, M., J. Eischeid, A. Kumar, R. Leung, A. Mariotti, K. Mo, et al. 2013a. Causes and predictability of the 2012 Great Plains drought. Bull. Am. Meteorol. Soc. 95:269–282. doi:10.1175/BAMS-D-13-00055.1 Hoerling, M., S. Schubert, K. Mo. A. AghaKouchak, H. Berbery, J. Dong, et al. 2013b. An interpretation of the origins of the 2012 Central Great Plains drought: Assessment report. NOAA Drought Task Force, Climate Program Office, Washington, DC. ftp://ftp.oar.noaa.gov/CPO/pdf/ mapp/reports/2012-Drought-Interpretation-final.web-041113.pdf (accessed 3 Feb. 2014). Illston, B.G., and J.B. Basara. 2003. Analysis of short-term droughts in Oklahoma. Eos Trans. AGU 84(17):157–164. doi:10.1029/2003EO170001 Intergovernmental Panel on Climate Change. 2000. Special report: Emissions scenarios. Cambridge Univ. Press, Cambridge, UK. Jones, J.W., J.W. Hansen, F.S. Royce, and C.D. Messina. 2000. Potential benefits of climate forecasting to agriculture. Agric. Ecosyst. Environ. 82:169–184. doi:10.1016/S0167-8809(00)00225-5 Karl, T.R., J.M. Melillo, and T.C. Peterson, editors. 2009. Global climate change impacts in the United States. Cambridge Univ. Press, Cambridge, UK. http://downloads.globalchange.gov/usimpacts/pdfs/climate-impactsrreports.pdf (accessed 3 Feb. 2014). Lal, R., J.A. Delgado, P.M. Groffman, N. Millar, C. Dell, and A. Rotz. 2011. Management to mitigate and adapt to climate change. J. Soil Water Conserv. 66:276–285. doi:10.2489/jswc.66.4.276 Li, L., W. Li, and Y. Deng. 2013. Summer rainfall variability over the southeastern United States and its intensification in the 21st century as assessed by CMIP5 models. J. Geophys. Res. Atmos. 118:340–354. doi:10.1002/ jgrd.50136

Li, W., L. Li, R. Fu, Y. Deng, and H. Wang. 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. 24:1449–1506. Liang, X., D.P. Lettenmaier, E.F. Wood, and S.J. Burges. 1994. A simple hydrologically based model of the land surface water and energy fluxes for general circulation models. J. Geophys. Res. 99(D7):14415–14428. doi:10.1029/94JD00483 Maurer, E.P., L. Brekke, T. Pruitt, and P.B. Buffy. 2007. Fine-resolution climate projections enhance regional climate change impact studies. Eos Trans. AGU 88(47):504. doi:10.1029/2007EO470006 McPherson, R.A., C.A. Fiebrich, K.C. Crawford, R.L. Elliott, J.R. Kilby, D.L. Grimsley, et al. 2007. Statewide monitoring of the mesoscale environment: A technical update on the Oklahoma Mesonet. J. Atmos. Ocean. Technol. 24:301–321. doi:10.1175/JTECH1976.1 Menne, M.J., C.N. Williams, and R.S. Vose. 2009. The United States Historical Climatology Network monthly temperature data, version 2. Bull. Am. Meteorol. Soc. 90:993–1007. doi:10.1175/2008BAMS2613.1 National Climate Assessment and Development Advisory Committee. 2009. Federal advisory committee draft climate assessment. U.S. Global Change Res. Prog., Washington, DC. http://ncadac.globalchange.gov/ (accessed 3 Feb. 2014) National Climatic Data Center. 2012. U.S. climate divisions. NCDC, Asheville, NC. http://www.ncdc.noaa.gov/temp-and-precip/us-climate-divisions. php (accessed 3 Feb. 2014). Nearing, M.A., F.F. Pruski, and M.R. O’Neal. 2004. Expected climate change impacts on soil erosion rates: A review. J. Soil Water Conserv. 59:43–50. Peterson, T.C., R.R. Heim, Jr., R. Hirsch, D.P. Kaiser, H. Brooks, N.S. Diffenbaugh, et al. 2013. Monitoring and understanding changes in heat waves, cold waves, floods and droughts in the United States. Bull. Am. Meteorol. Soc. 94:821–834. Starks, P.J., C.A. Fiebrich, D.L. Grimsley, J.D. Garbrecht, J.S. Steiner, J.A. Guzman, and D.N. Moriasi. 2014. Upper Washita River experimental watersheds: Meteorologic and soil climate measurement networks. J. Environ. Qual. doi:10.2134/jeq2013.08.0312 Thiessen, A.H. 1911. Precipitation for large areas. Mon. Weather Rev. 39:1082–1084.

www.agronomy.org • www.crops.org • www.soils.org 1327

Climate change and observed climate trends in the fort cobb experimental watershed.

Recurring droughts in the Southern Great Plains of the United States are stressing the landscape, increasing uncertainty and risk in agricultural prod...
4MB Sizes 0 Downloads 7 Views