1. Introduction and background
The United States is the largest producer and exporter of corn in the world, and corn is the most widely produced feed grain in the United States (accounting for more than 95% of total production and use; Economic Research Service 2020). With a projected global population approaching $9.7 billion by 2050 (United Nations 2019), society is faced with the challenge of keeping food production on pace with global population growth. In addition to increasing demand, producers are also faced with potential impacts from anthropogenic climate change in the form of increased exposure to warmer than normal temperatures, prolonged drought conditions, and more extreme precipitation events (Nicholls 1996; Karl and Knight 1998; Milly et al. 2002; IPCC 2007, 2013, 2021; Bengtsson 2010; Walthall et al. 2013; Pryor et al. 2014; Lesk et al. 2016; Mase et al. 2017; Steele and Hatfield 2018; Steiner et al. 2018). The aggregate societal impacts from weather and climate extremes, and trends in those impacts, are a function of weather, climate, and society (Changnon et al. 2000; S. A. Changnon 2003; S. D. Changnon; Hallegatte et al. 2007). For corn, production has increased nearly eightfold since 1940 (National Agricultural Statistics Service 2021), and this logically relates to the expanding bull’s-eye concept for corn producers. That is, “targets”—humans and their possessions—of geophysical hazards are enlarging as populations grow and spread (Ashley et al. 2014; Strader and Ashley 2015). Increased exposure to prolonged adverse conditions can cause shifts in production efficiency and drive crop losses that threaten food security and economic well-being (Schlenker and Roberts 2009; Pryor et al. 2014; Elias et al. 2018; Kistner et al. 2018; Steele and Hatfield 2018; Steiner et al. 2018).
To mitigate the potential negative impacts of weather and climate to agriculture, producers often obtain crop insurance, which is an important risk management tool by providing a financial safety net (Walthall et al. 2013; Reyes et al. 2020). Crop insurance is a financially stable risk management tactic and provides opportunities for farmers to make long-term investments to adapt to the changes that climate change may bring to agronomic conditions (Mieno et al. 2018). The U.S. Department of Agriculture (USDA) Risk Management Agency (RMA) administers the federal crop insurance program and offers multiple insurance plans and coverage levels to mitigate revenue losses, yield losses, and damage to crops by weather and other perils (Shields 2015). Producers receive an indemnity payment when crops are damaged or lost due to insurable perils. These indemnity payments are based on the insureds’ coverage level, liabilities, specific program policies, and the extent and magnitude of the damage (RMA 2018). Perils causing crop loss, which are established by producers and verified by claim adjusters from the RMA or private insurance companies, are often due to natural hazards. Historical crop indemnity data based on weather and climate-driven losses can offer insights on both the biophysical and socioeconomic vulnerabilities of agriculture (Reyes and Elias 2019). In addition, these data can be used to examine trends over time and assess impacts of past weather and climate-driven events on agricultural production (Changnon et al. 2000; Rosenzweig et al. 2002; Lobell et al. 2011; Smith and Katz 2013; Smith and Matthews 2015; Rohli et al. 2016; Reyes and Elias 2019; Reyes et al. 2020). Since 1989, social and legislative changes have affected the federal crop insurance program resulting in increased losses. This includes the 2000 Agricultural Risk Protection Act; farm bills in 1996, 2002, 2008, and 2014; increases in premium subsidies since 1994; and inflation (Congressional Budget Office 2017). Thus, socioeconomic change is the primary factor responsible for the upward trend in loss, and many studies agree that this will continue as the main factor driving loss throughout the twenty-first century (Kunkel et al. 1999; Pielke 2005; Höppe and Pielke 2006; Bouwer 2011; Barthel and Neumayer 2012; IPCC 2012a; Strader et al. 2017). However, climate change may amplify this risk for many perils (e.g., Strader and Ashley 2015). Therefore, normalized crop loss is key for understanding the impacts of interannual changes in crop prices, RMA crop insurance policies, and socioeconomic conditions (Changnon et al. 2000; Changnon and Hewings 2001; Barthel and Neumayer 2012; Smith and Katz 2013). Additional accounting for trends in weather and climate perils is useful and is considered nontrivial as well (Changnon and Hewings 2001).
Assessments of historical corn losses in the form of insurance payments offers valuable insight that can be implemented into risk management strategies, especially when used in the context of climate change (Kistner et al. 2018; Steele and Hatfield 2018; Reyes and Elias 2019; Reyes et al. 2020). Given this, the primary objective of this study was to establish a spatiotemporal weather and climate peril climatology for metrics related to corn indemnity. Metrics examined include corn payment indemnity (CPI)-adjusted indemnity, cost loss, and acreage loss. Quantifying spatiotemporal trends for corn losses at county-level spatial resolution for the United States using indemnity metrics was a key objective to this study, which has not been performed to date. This study also builds on previous works that have used insurance as a proxy for agricultural impacts by distinct weather and climate-related hazards (Reyes and Elias 2019; Reyes et al. 2020) by 1) performing analyses at county-level spatial resolution to highlight counties utilizing corn insurance as an adaptation strategy against certain perils, 2) investigating a longer study period, and 3) examining three distinct indemnity metrics. This study also helps to inform risk management decisions by highlighting high risk corn production areas and furthers the understanding of agricultural vulnerability (Wallander et al. 2013; Reyes et al. 2020). Results herein also inform corn crop peril risk mitigation strategies [e.g., shifting production systems, increasing crop insurance coverage, and increasing the number of acres (1 acre = 0.4 ha) insured], which will help to inform a more resilient and sustainable U.S. corn production system.
2. Data and methods
Corn insurance data were obtained through the USDA RMA Cause of Loss and Summary of Business datasets (RMA 2021) for the period 1989–2020. Cause of Loss and Summary of Business data are publicly available dating back to 1989, which informs the 32-yr dataset for this analysis. Several weather and climate causes of loss, including water deficit, water surplus, severe weather, and cool-season perils were examined. Specifically, drought and heat were classified as water deficit perils, excess moisture/rain/precipitation (excess moisture hereinafter) and flood were classified as water surplus perils, hail and damaging wind were noted as severe weather perils, and cold wet weather and freeze/frost were labeled as cool-season perils. Indemnities, and the number of acres damaged per event, were aggregated monthly and annually at the county level. RMA Cause of Loss data were merged with the RMA Summary of Business dataset, which contains county-level liabilities and number of acres insured. The two datasets were merged based monthly sum of liabilities and the monthly sum of payment indemnity by peril for each county and peril.
Analyses contained three distinct metrics for analyzing corn insurance. The first used CPA-inflation adjusted for 2020 USD payment indemnity (i.e., amount paid to the producer by the insurer due to a loss based on insurance coverage and loss claim) to quantify payout averages and trends over time. Indemnity data were also linearly detrended to eliminate any existing long-term trends associated with each peril. Standardized anomalies were then calculated using detrended indemnity data to assess the long-term interannual variability for each peril at the U.S. aggregated level. While there have been programmatic and policy changes within the 32-yr study that influence long-term trends in corn loss, these effects are minimized by calculating monthly and annual county-level loss cost. Loss cost presents a significant advantage for this type of analysis, as it accounts for interannual changes in specific commodity prices, RMA program policies, and socioeconomic conditions (Changnon and Hewings 2001; Barthel and Neumayer 2012). In other words, loss cost is not affected by inflation or price change over time (Li et al. 2019). Loss cost has been extensively used for examining crop insurance losses through time (Changnon and Changnon 1997, 2000; Coble and Barnett 2013; Reyes and Elias 2019; Reyes et al. 2020; Perry et al. 2020). This unitless value is calculated as (indemnity/liability) × 100 for the predetermined peril, time interval, and spatial resolution. Essentially, loss cost is a normalized value for indemnities accounting for the value of products (Reyes and Elias 2019). Assessing crop losses relative to liabilities normalizes indemnities by the current year’s aggregated commodity prices and is an efficient and effective way to analyze trends in agricultural losses (Changnon et al. 2000). While loss cost is a simple metric when analyzing agricultural loss, it overcomes the challenge of commodity prices and inflation impacting indemnity data. The third metric assessed was corn acreage loss percentage, calculated by finding the acres lost at monthly and annual intervals with respect to the number of acres insured as a percentage. This is calculated as (acres damaged/acres insured) × 100 for the predetermined peril, time interval, and spatial resolution. This acreage based normalization aims to assess trends in the number of acres damaged by a peril. Over the past 25 years, the federal crop insurance program has expanded with substantial increases in insured acres, subsidies, and liability (Glauber 2013), all of which impact the trend in indemnity and the trend in acreage damaged by a peril. Thus, normalization techniques such as loss cost and acreage loss were vital to examine.
Trend analyses were performed on inflation-adjusted indemnities, loss cost, and acreage loss using Theil–Sen’s slope due to its efficient computation and insensitivity to outliers (Wilcox 2010). Statistical significance of Theil–Sen’s slope was assessed using Kendall’s τ statistic and a p value at the 0.05 (95% confidence) significance level. Trends were assessed at monthly and annual intervals for each peril over the 32-yr period. Spatial analysis was limited to counties east of the 105° west meridian, which accounts for 99% of U.S. corn production (National Agricultural Statistics Service 2019). This domain was also stratified into USDA Farm Production Regions and will be used as guidance when discussing results herein (Fig. 1; Cooter et al. 2012).
Study area and USDA farm production regions.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Study area and USDA farm production regions.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Study area and USDA farm production regions.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Spatially, trends were quantified at the county level. Analyses were also restricted to counties based on number of payments that varied by peril due to interannual variability. If 32 years of data for each county in order to calculate averages or trends were required, then drought and excess moisture would likely be the only two qualifying perils based on previous studies (Reyes et al. 2020; Perry et al. 2020). The following are the minimum number of years with an indemnity payment required by peril for a county to be included in the spatial analysis: drought: 15, heat: 10, excess moisture: 15, flood: 5, cold wet weather: 10, freeze/frost: 5, hail: 10, excess wind: 10. These were determined by taking the average number of years each county had for each peril and rounding that average number to the nearest 5. For example, the average number of years each county had losses due to drought was 15 throughout the United States; therefore, any county that had less than 15 years of drought loss were not used in the spatial analysis. This restriction aimed to keep the results robust for each peril (i.e., mitigate issues due to lack of sample size).
3. Results
a. Climatology
1) Monthly climatology
(i) Water deficit perils
Nearly 2000 U.S. counties reported at least one weather or climate loss to corn during the 32-yr study period 1989–2000. Almost 800 000 total payments were made to corn producers, resulting in a cumulative payment indemnity of $43 billion. When examining monthly loss cost and acreage loss, a strong correlation existed (R = 0.90). However, between inflation-adjusted indemnity and loss cost/acreage loss, the correlation was noted as weak (R = 0.32). Nonetheless, using these insurance metrics to evaluate corn loss will provide unique insight into peril risk (i.e., provide a climatological overview). Over the 32-yr study period, drought and excess moisture were the two costliest and most frequent perils for corn in the United States, accounting for about 84% of all weather and climate-driven indemnities and 72% of total payments made, underscoring the importance of water availability in corn production. Drought and heat, categorized here as water deficit perils, accounted for 47% and 5% of the 32-yr indemnity total, respectively. These two perils were generally concurrent with each other and peaked in indemnity payout during July (Figs. 2a,b). July is also characterized by the temperature-sensitive corn pollination stage, and has major implications on the physiology, quality, and production of corn in any given season (Lonnquist and Jugenheimer 1943; Herrero and Johnson 1980; Bundy and Gensini 2022). July 2012 was the costliest month for these two perils (U.S. indemnity surpassed $10 billion for corn) because of the historical drought, which encompassed many of the top corn-producing states and led to one of the top 5 hottest/driest summers over the last 127 years (NCEI 2021). This event accounted for almost one-quarter of the total 32-yr weather and climate-driven indemnity total.
The 1989–2020 total sums of U.S. payment indemnity (blue; left axis; millions of USD), loss cost divided by 500 (black; right axis; unitless), and acreage loss divided by 1000 (red; right axis; unitless) by month and weather/climate peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
The 1989–2020 total sums of U.S. payment indemnity (blue; left axis; millions of USD), loss cost divided by 500 (black; right axis; unitless), and acreage loss divided by 1000 (red; right axis; unitless) by month and weather/climate peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
The 1989–2020 total sums of U.S. payment indemnity (blue; left axis; millions of USD), loss cost divided by 500 (black; right axis; unitless), and acreage loss divided by 1000 (red; right axis; unitless) by month and weather/climate peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
(ii) Water surplus perils
Conversely, perils that led to water surplus (excess moisture and flood) accounted for nearly 36% and 2% of the 32-yr indemnity total, respectively. These perils were costliest during the early corn growing season (April–June), then decreased in indemnity payout totals during the latter half of the season (Figs. 2c,d). Losses due to excess moisture totaled nearly half of all loss payments during spring months, which is consistent with previous research (Reyes and Elias 2019). May 2019 was the costliest month for excess moisture and flood induced corn losses, with the U.S. accumulated indemnity totaling $1.3 billion, as many top corn-producing states recorded precipitation totals ranking in the top 10 of the last 127 years (NCEI 2021). In addition, Palmer drought severity index values were ranked as the wettest in the historical record across a majority of the Corn Belt (NCEI 2021), causing many fields to be inundated and unfavorable for planting and fieldwork.
(iii) Severe weather perils
Hail was the costliest severe weather peril (i.e., hail, excess winds, and tornadoes) for corn. Hail accounted for 6% of all indemnities for corn, and when combined with damaging winds, accounted for 8% of the weather and climate-driven peril 32-yr indemnity total (tornado losses were not used due to very limited observations). Hail was the costliest during June and July (Fig. 2e), which coincides with the peak U.S. hail occurrence (Cintineo et al. 2012), and when the susceptibility of corn to hail damage is high (Changnon 1971; Changnon and Changnon 1997). June 2014 was the costliest month in the study period for corn loss due to hail, as the U.S. accumulated indemnity totaled $247 million. Like hail, damaging winds tended to have the greatest impact when corn is at its most vulnerable stage, with taller stalks being more susceptible to blow down (Cleugh et al. 1998). The largest indemnities and losses occurred in July and August (Fig. 1f), which coincides with peak derecho frequency across the Corn Belt (Ashley and Mote 2005; Guastini and Bosart 2016). The costliest month in the study period for severe wind driven corn loss was August 2020 (direct result of the 10 August 2020 derecho event across the heart of the Corn Belt), with an accumulated U.S. indemnity exceeding $261 million.
(iv) Cool-season perils
Cool-season perils (cold wet weather and freeze/frost), while not as frequent, still accounted for 3% of the total weather and climate-driven indemnity. Even though historical indemnities were relatively small in comparison with other perils, there is still a risk for corn to be affected by cold wet weather and freeze/frost. In large part, this is due to improved equipment efficiency and the development of hybrids with a greater tolerance to suboptimal conditions, leading to large-scale shifts toward earlier planting over the last three decades (Bruns and Abbas 2006; Lauer 2001; Westcott and Jewison 2013). Freeze/frost was most costly during harvest time (September–October), but early months (April–May) still proved costly (Fig. 2g). September 1993 was the costliest month for freeze/frost, as many of the top corn-producing states recorded monthly temperatures rankings in the top 5 coldest for September in the last 127 years (NCEI 2021). The aforementioned May 2019 was the costliest month for cold wet weather in this study period. Early months, such as April and May, tended to have the highest frequencies of cold wet weather losses (Fig. 2h), following a similar pattern to excess moisture (Fig. 2c).
2) Annual county-level spatial climatology
(i) Water deficit perils
Drought was the costliest peril in terms of total indemnity accumulation since 1989 in much of the southern plains, Corn Belt, Southeast, and Appalachian regions (Fig. 3). In terms of annual averages, the largest for corn payment indemnity due to drought were confined to the Corn Belt, northern plains, and a small portion of the Mountain region (Fig. 4a), with relative maxima noted in eastern Illinois, northwest Iowa, South Dakota, and eastern Colorado reaching $3.5 million yr−1. The general pattern of high drought indemnity values displayed a similar spatial distribution as to where major corn acreage is located, which will be further explained in the discussion section. Examining loss cost and acreage loss, the plains regions, Southeast, and Appalachian regions exhibited the highest annual averages (Figs. 4b,c). Highly irrigated areas, such as south-central Nebraska and southwest Kansas, unsurprisingly recorded lower drought indemnity averages. Drought indemnities were highest in areas where the majority of fields are rainfed (e.g., Illinois, Iowa), thus, more susceptible to rainfall variability (Li et al. 2019). No drought loss data were displayed in the Texas Panhandle, nor in the delta region, because of high irrigation rates that reduced the number of years with drought payment indemnity to under the 15-yr threshold. Therefore, in the Texas/Oklahoma Panhandles and southwestern Kansas, heat has been the costliest peril due to regional dominance of irrigation (Fig. 3). It is notable, however, that corn losses due to drought are distinguished in the RMA data from irrigated crops that do not receive sufficient water, as these losses are reported as failure of irrigation supply. Failure of irrigation supply refers to the lack of physical water availability for irrigation and not due to failure of irrigation equipment, which has been quantified by previous literature to be a key driver in crop loss in the southern plains (Reyes et al. 2020).
Costliest weather/climate peril by county as based on the 1989–2020 sum of payment indemnity. The top 50 corn-producing counties are outlined in black.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Costliest weather/climate peril by county as based on the 1989–2020 sum of payment indemnity. The top 50 corn-producing counties are outlined in black.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Costliest weather/climate peril by county as based on the 1989–2020 sum of payment indemnity. The top 50 corn-producing counties are outlined in black.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Annual averages for corn payment indemnity (USD), loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus perils (1989–2020). Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥15 years for drought; ≥10 years for heat; ≥15 years for excess moisture; ≥5 years for flood).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Annual averages for corn payment indemnity (USD), loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus perils (1989–2020). Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥15 years for drought; ≥10 years for heat; ≥15 years for excess moisture; ≥5 years for flood).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Annual averages for corn payment indemnity (USD), loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus perils (1989–2020). Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥15 years for drought; ≥10 years for heat; ≥15 years for excess moisture; ≥5 years for flood).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Losses due to heat were the highest in the southern and eastern mountain region as well as the southern plains, collocating with where the risk for the warmest temperatures is climatologically the greatest during the growing season (Fig. 4d). Specifically, this includes counties in the Texas and Oklahoma Panhandles, western Kansas, and eastern Colorado, with averages totaling over $670,000 yr−1. Like drought, areas outside major production acreage (including the western domain of the northern plains, much of the southern plains, mountain, delta, and eastern U.S. regions) have the highest annual averages of loss cost (>15) and acreage loss (>21; Figs. 4e,f). Higher loss cost and acreage loss here suggests that indemnity payouts are much closer to the total liability on average, and the number of acres reported as damaged is much closer to the number of acres insured on average.
(ii) Water surplus perils
The highest annual averages for water surplus perils were confined to acreage in the northern plains and along major rivers (Figs. 4g,j). North Dakota, South Dakota, Minnesota, and acreage along the Missouri River, Red River, Mississippi River, and Ohio River possessed a higher risk for these perils when averaged out over the 32-yr period. This does not necessarily follow what the annual precipitation climatology or heavy rainfall climatology presents, however (Brooks and Stensrud 2000; NCEI 2021), suggesting there is regional variability in terms of reaction to excess precipitation events that induce water surplus scenarios that will be further explored in the discussion section below. Both North and South Dakota recorded counties with an annual average indemnity payment of greater than $4.7 million yr−1 for excess moisture. It is clear that excess moisture and the aforementioned drought peril were the two costliest and most widespread, accounting for the maximum accumulated loss in 1890 of the 2084 (91%) counties producing corn (Fig. 3). The indemnity hotspot for flooding was located along the Missouri River in far eastern Nebraska with annual averages near $2.2 million yr−1. For loss cost and acreage loss, some of the largest values were located in northern acreage including South Dakota and Minnesota for excessive moisture, and also located in Corn Belt acreage including Missouri and the southern delta region for both excess moisture and flood (Figs. 4h,i,k,l).
(iii) Severe weather perils
Relative maxima for corn indemnity severe weather annual averages were located in the western Corn Belt and throughout the Great Plains (both northern and southern plains) regions, including counties in Texas, Kansas, Colorado, Nebraska, and Iowa (Figs. 5a,d). For hail specifically, northeastern Colorado and southwestern Nebraska exhibited the highest indemnity payments throughout the country, with annual averages totaling $1.5 million yr−1. This collocates with where the most hail reports, where the most large-hail environments and reports and severe hail days are located during the peak growing season (Schaefer et al. 2004; Cintineo et al. 2012; Allen et al. 2015). Further, hail has been the costliest peril in western portions of Nebraska and in numerous counties in the Great Plains and mountain regions (Fig. 3). Loss cost and acreage loss tended to have the highest annual averages focused along the western extent of the Great Plains (Figs. 5b,c). Portions of central and eastern Iowa recorded the highest annual average indemnity due to excess wind, which was largely driven by the costliest thunderstorm in U.S. history—the 10 August 2020 Midwest derecho (NCEI 2021). As a result of this event, Marshall and Tama Counties in Iowa now possess the highest annual average loss cost and acreage loss by excessive wind in the United States, and excessive wind is also the costliest peril for these counties (Figs. 5e,f and 3). Drought was the costliest peril in these counties prior to the 2020 derecho, and excessive wind was the highest (on average) across eastern Colorado and western Nebraska.
As in Fig. 4, but for (a)–(f) severe weather and (g)–(l) cool-season perils. Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥10 years for hail and excess wind; ≥10 for cold wet weather; ≥5 for freeze/frost).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 4, but for (a)–(f) severe weather and (g)–(l) cool-season perils. Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥10 years for hail and excess wind; ≥10 for cold wet weather; ≥5 for freeze/frost).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 4, but for (a)–(f) severe weather and (g)–(l) cool-season perils. Counties shaded in gray represent corn-producing counties that did not meet the sample size criteria (≥10 years for hail and excess wind; ≥10 for cold wet weather; ≥5 for freeze/frost).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
(iv) Cool-season perils
The cool-season perils (cold wet weather and freeze/frost) have also been biased to northern corn acreage in the northern plains and lake regions with higher indemnity averages and more counties meeting the threshold (≥10 years for cold wet weather; ≥5 years for freeze/frost), unsurprisingly, because of colder annual average temperatures (Figs. 5g–l). Nebraska, the Dakotas, Minnesota, and Wisconsin recorded the highest indemnities, highest loss cost, and highest acreage loss for both cold wet weather and freeze/frost, signifying higher risk in these areas to these perils. For these cool-season perils, 10 of the 12 counties with freeze/frost or cold wet weather as the costliest perils are confined to the northern most counties in North Dakota, Minnesota, Wisconsin, and Michigan (Fig. 3). Overall, the combination of excess moisture and cold wet conditions both peaking in the early growing season creates a dual-threat peril for the majority of counties in these climatologically cooler regions.
b. Trends
1) Interannual trends
Within the 32-yr study period, inflation-adjusted corn payment indemnity for combined weather and climate-driven perils in the United States has increased by 374% (Fig. 6). The extremely anomalous years of 2012 (historical drought) and 2019 (extreme early season precipitation) were most noteworthy. Statistically significant increases in annual indemnities ($67 million yr−1) are attributed to the historically hot and dry period between 2011 and 2012. Furthermore, prior to 2013, loss cost for all weather and climate-driven perils was trending upward at 370 yr−1. The marginal increasing trend from 2013 onward (132 yr−1) may indicate higher production risk illustrated by much larger liabilities relative to indemnities, which suggests more hedging with higher premia paid for insurance coverage (Reyes and Elias 2019; Reyes et al. 2020). Liabilities increased by 754% ($1.5 billion yr−1) over the study period, which relates to socioeconomic factors (e.g., major pieces of legislation that affected the federal crop insurance program) and increased frequency of extreme weather and climate-related events. Congressional legislation affecting the federal crop insurance program (occurred in the years 1990, 1994, 2000, 2002, 2008, and 2014) (Congressional Budget Office 2017; Rosa 2018) has ultimately impacted insurance participation rates and total payouts. Annual acreage loss values have increased in the United States by 412 yr−1 (Fig. 6). Like loss cost, the increasing trend in acreage loss has also slowed after 2012 (338 yr−1), but not as significantly as loss cost trends. Essentially, starting in 2013, more corn acreage was being insured and higher premiums were being paid in order to increase liability to hedge against potential losses by certain perils. The speculation of hedging occurring is only true if loss cost has decreased throughout time Reyes et al. (2020).
Interannual variability (1989–2020) of U.S. corn payment indemnity (CPI-adjusted billions of USD; right axis), loss cost, and acreage loss (unitless; left axis).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Interannual variability (1989–2020) of U.S. corn payment indemnity (CPI-adjusted billions of USD; right axis), loss cost, and acreage loss (unitless; left axis).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Interannual variability (1989–2020) of U.S. corn payment indemnity (CPI-adjusted billions of USD; right axis), loss cost, and acreage loss (unitless; left axis).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Marked increases in indemnities for multiple commodities in the United States disappear when using liabilities to calculate annual loss cost (Smith and Katz 2013). Similar to the findings of (Reyes and Elias 2019), results herein indicate that trends do exist, but are highly dependent on the peril and spatiotemporal resolution of aggregation. In addition, results here differ from the findings of Reyes and Elias (2019), as the single crop analysis provides more specific commodity risk insight at the county level. In other words, trends in agricultural loss are not only dependent on peril and spatiotemporal resolution, but also dependent on crop.
(i) Water deficit perils
Every peril examined, with the exception of freeze/frost, displayed a statistically significant increase in inflation-adjusted annual payment indemnities for corn in the United States (Figs. 7a,d,g,j, and 8a,d,g,j). These trends alone cannot explain changes in peril frequency. However, analyzing all three metrics, in addition to other research that has quantified trends in weather/climate hazards, provides further evidence to help explain the changing corn loss landscape. For example, drought has subtly decreased in loss cost but has subtly increased in acreage loss, concluding that there has not been a long-term trend over the study period (Figs. 7b,c). An anomalous decrease in drought loss cost occurred after 2012. As mentioned previously, when all perils were aggregated together and categorized as weather and climate-driven losses, loss cost decreases after this period as well, suggesting that drought played a substantial role in the overall trend in loss cost after 2012. Loss cost for all weather-climate driven losses combined increased by 132 yr−1 from 2013 to 2020, which was significantly less than 1989–2012 trend (369 yr−1). For drought, loss cost increased by only 85 yr−1 from 2013 to 2020, which represented the largest decline in loss cost trend between the 1989–2012 and 2013–20 epochs. The same pattern of loss cost displayed can be recognized in acreage loss after the year 2012. For heat, which is generally coupled with drought, loss cost and acreage loss have both increased, with the increasing trend in acreage loss being statistically significant (Figs. 7e,f). No significant trends in loss cost for drought were noted over 32-yr study period, which is similar to the findings in Reyes and Elias (2019) when assessing all agriculture. Any differential trends in loss cost between perils underscores the importance of spatial and temporal resolution, and suggests further analysis is necessary when using crop-loss data for decision-making (Reyes and Elias 2019). In addition, any differentiating trends between literature, such as this one and Reyes and Elias (2019), highlights the importance of crop selection when analyzing these data.
The 1989–2020 variability of U.S. corn payment indemnity annual totals, loss cost annual totals, and acreage loss annual totals for (a)–(f) water deficit and (g)–(l) water surplus perils. Theil–Sen slopes are plotted and shown as solid lines if significant at the 95% confidence level using Kendall’s τ test. Slope values are given in boldface type if the trend was statistically significant.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
The 1989–2020 variability of U.S. corn payment indemnity annual totals, loss cost annual totals, and acreage loss annual totals for (a)–(f) water deficit and (g)–(l) water surplus perils. Theil–Sen slopes are plotted and shown as solid lines if significant at the 95% confidence level using Kendall’s τ test. Slope values are given in boldface type if the trend was statistically significant.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
The 1989–2020 variability of U.S. corn payment indemnity annual totals, loss cost annual totals, and acreage loss annual totals for (a)–(f) water deficit and (g)–(l) water surplus perils. Theil–Sen slopes are plotted and shown as solid lines if significant at the 95% confidence level using Kendall’s τ test. Slope values are given in boldface type if the trend was statistically significant.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 7, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 7, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 7, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
(ii) Water surplus perils
The largest peril increases in indemnity resulted from excess moisture. The $26.1 million yr−1 increase was more than double the second highest increasing trend (drought at $12.1 million yr−1). The statistically significant increases found in payment indemnity, loss cost, and acreage loss (Figs. 7g–i) occurred simultaneously with a significant increase in peril frequency (e.g., Kunkel 2003; Berardi et al. 2019; Changnon and Gensini 2019). Substantial increases in loss cost for the period replicated the increase in loss cost drought displayed prior to the 2012 historical drought. This trend is substantially different than what was quantified in Reyes and Elias (2019), thus, further demonstrating that loss trends are also dependent on the crop selection. Unlike excess moisture, the flood peril for corn has decreased in loss cost and acreage loss, albeit not statistically significantly, likely due to flooding being more on the extreme end of excessive water conditions and thus promoting the need for utilizing crop insurance (Figs. 7k,l).
(iii) Severe weather perils
Hail and excessive wind have both displayed statistically significant increases in indemnities over the study period (Figs. 8a,d). The increase in crop insurance coverage is warranted given the increase in hail events over the past 100 years (Changnon and Changnon 2000), especially in Nebraska where a large percentage of indemnities for hail are located. However, loss cost and acreage loss did not display statistically significant trends for hail loss. Conversely, corn losses due to excess wind have increased significantly in both loss cost and acreage loss (Figs. 8e,f). This peril may not receive as much attention due to the scale in which the events occur on; however, it is evident that with a statistically significant increase in indemnity, loss cost, and acreage loss, that the frequency or intensity of this peril has possibly increased as well and should be considered when evaluating future risk to corn production.
(iv) Cool-season perils
Cold wet weather and freeze/frost had differential trends in indemnity, loss cost, and acreage loss. Cold wet weather has increased in all three metrics, but only statistically significantly for indemnity (Figs. 8g–i). Freeze/frost decreased in all three metrics with loss cost and acreage loss being statistically significant (Figs. 8j–l). At the national scale, loss cost trends differed from increasing loss cost trends quantified by Reyes and Elias (2019), which, again, examined total crop loss rather than just corn. Since peak cold wet weather risk typically occurs at the same time as peak excess moisture within the growing season, it is reasonable to speculate that trends in cold wet weather will mimic trends in excess moisture. With the growing season starting earlier in recent decades, exposure to cold wet weather has likely increased, thus increasing indemnities that fall under this peril. Though, these increasing trends for loss cost and acreage loss were deemed insignificant. Under a warming climate, the overall decrease in freeze/frost indemnity, and statistically significant decreases in loss cost and acreage loss could be attributed to increasing daily minimum temperatures (including the decrease of record-setting cold temperatures; Vose et al. 2005; Meehl et al. 2009), the advancements in corn hybrids, enhancement of other mitigation technologies (Cooper et al. 2014), or to other reasons such as hedging.
(v) Trend-adjusted indemnities
When considering socioeconomic factors (e.g., increases in acres planted), corn losses trend upward. An alternative way to examine this is to detrend the indemnity data and quantify the standardized anomalies σ for each peril on an annual basis (Fig. 9). This analysis indicated that all perils, with the exception of freeze/frost, have more years with indemnity payments above 1σ in the second half of the study period. For water deficit perils, 2011 and 2012 were the only years above 1σ. Years above 1σ for water abundance perils included 1993, 2008, 2011, 2013, and 2019. Severe weather years above 1σ included 2009, 2011, 2014, and 2020. Most perils also had a number of years below −1σ in the second half of the study period, suggesting that while indemnities have increased overall for all perils except freeze/frost, the interannual variability in these perils has also increased over the last two decades, which has been noted with many weather and climate extremes (IPCC 2012b; Thornton et al. 2014).
Interannual heat map of trend-adjusted payment indemnity standardized anomalies by weather peril (1989–2020).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Interannual heat map of trend-adjusted payment indemnity standardized anomalies by weather peril (1989–2020).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Interannual heat map of trend-adjusted payment indemnity standardized anomalies by weather peril (1989–2020).
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
2) Spatial trends
(i) Water deficit perils
The most robust increases in drought indemnity were located in portions of Kansas, South Dakota, North Dakota, Minnesota, and Wisconsin, ranging between $10,000–126,000 yr−1 (Fig. 10a). These areas have been associated with some of the most severe droughts in U.S. history (Ganguli and Ganguly 2016). A decreasing trend in eastern Nebraska showed that indemnity on an annual basis for corn had a marginal decrease over the study period, suggesting that irrigation methods have likely prevented insurance claims due to drought from increasing, or that any losses due to “lack of water” are reported as failure of irrigation supply. Eighty-one percent of qualifying corn-producing counties had an increase in drought payment indemnity, about a fifth of them being statistically significant. Loss cost and acreage loss had the most robust decreasing trends in the northern plains across Nebraska, South Dakota, and North Dakota (Figs. 10b,c). Further examination into drought may be warranted, especially before and after the 2012 drought, as this historical event played a major role in the overall 32-yr indemnity, loss cost, and acreage loss trends at the national level.
Theil–Sen slopes of the annual payment indemnity, loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus peril (1989–2020). Counties outlined in black signify statistical significance at the 95% confidence level using Kendall’s τ test. Counties shaded in gray did not meet the sample size criterion for the respective peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Theil–Sen slopes of the annual payment indemnity, loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus peril (1989–2020). Counties outlined in black signify statistical significance at the 95% confidence level using Kendall’s τ test. Counties shaded in gray did not meet the sample size criterion for the respective peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Theil–Sen slopes of the annual payment indemnity, loss cost, and acreage loss by (a)–(f) water deficit and (g)–(l) water surplus peril (1989–2020). Counties outlined in black signify statistical significance at the 95% confidence level using Kendall’s τ test. Counties shaded in gray did not meet the sample size criterion for the respective peril.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Spatially, heat displayed scattered statistically significant upward trends in indemnity throughout the study domain (Fig. 10d). Many counties with statistically significant increases were also within the multiple clusters of counties with significant increases in drought payment indemnity; 75% of qualifying U.S. corn-producing counties had an increase in heat payment indemnity, with 11% of those counties being statistically significant at the 95% confidence level. One cluster of significant increases in heat is located in the Texas Panhandle, suggesting that this area still corresponds to a high-risk production area (with potential for prolonged droughts, excessive high temperatures, increasing water scarcity, and depleting groundwater levels) due to this area being reliant on irrigation (Cayan et al. 2010; Scanlon et al. 2012; Cook et al. 2015; Reyes and Elias 2019; Reyes et al. 2020). This area also experienced the most robust increases in loss cost and acreage loss in the study domain (Figs. 10e,f). Another notable area where robust increases in indemnity, loss cost, and acreage loss have occurred was in the Appalachian and Southeast regions. Like drought, significant decreases have also occurred in heat loss cost and acreage loss in the northern plains.
(ii) Water surplus perils
The most widespread and robust of the increases in payment indemnity for corn in the United States have been due to excessive moisture. Increasing heavy rainfall during the growing season has increased in recent decades (Angel et al. 2018; Changnon and Gensini 2019) and is projected to occur more frequently in the future (Kunkel 2003; Berardi et al. 2019). The most robust increases in excess moisture have occurred along the Mississippi River valley in the delta region, the eastern Midwest, and in the northwest domain of the Midwest, with the highest increases nearing $400,000 yr−1 (Fig. 10g). This collocates with states that have experienced robust increases in annual precipitation, increases in 99th-percentile precipitation value, and number of 5-yr, 2-day precipitation events over the past 20 years (Easterling et al. 2017). Ninety-seven percent of qualifying corn-producing counties in the United States have had an increase in excessive moisture payment indemnity, 59% of which were statistically significant. Excess moisture loss cost and acreage loss also displayed the most more widespread increases in acreage outside the Corn Belt (Figs. 10h,i). The delta region, in particular, has been a hotspot for the highest increases in loss cost and acreage loss, suggesting this region may be deemed riskier than what previous research has discussed for this peril (e.g., Reyes and Elias 2019; Perry et al. 2020). Thus, this is a location that should be further examined, perhaps at a higher temporal resolution and for other crops to investigate monthly and seasonal risk for various producers in the region.
Flooding indemnity increases were found throughout many of the Corn Belt counties, and with maximum increases at $45,000 yr−1 located along the Mississippi River in the far southern delta region (Fig. 10j). Other localized maxima flagged as statistically significant were found along the Mississippi River in Missouri and along the Missouri River in Missouri and Iowa. However, loss cost and acreage loss due to flooding have both decreased throughout much of the Corn Belt (Figs. 10k,l). As mentioned earlier, flooding tends to be on the extreme end of excess moisture, and thus, would be more likely to trigger a decline in yield potential, further promoting the need for insurance.
(iii) Severe weather perils
The largest increases in hail induced corn payment indemnity coincided with the highest annual average payment indemnity areas, ranging from increases of $40,000–95,000 yr−1 in the Texas and Oklahoma Panhandles, western Kansas, eastern Colorado, and western Nebraska (Fig. 11a). Such a substantial increase in this area is likely attributed to the increase in acres being planted (i.e., the agricultural expanding bull’s-eye effect; see section 4) and the increase in large-hail reports and environments over the last 40 years (Tang et al. 2019; Gensini and Brooks 2018; Gensini et al. 2020). Eighty-four percent of qualifying U.S. corn-producing counties experienced an increase in payment indemnity for hail induced corn loss, and 32% of those increases were statistically significant. Widespread decreases of loss cost for hail have been observed, especially in eastern Nebraska and Kansas and through the Dakotas (Fig. 11b). As for acreage loss, the most robust increases have been observed in western Nebraska (Fig. 11c).
As in Fig. 10, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 10, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
As in Fig. 10, but for (a)–(f) severe weather and (g)–(l) cool-season perils.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Trends in payment indemnity due to excess wind mimic the spatial patterns for hail trends (Fig. 11d). Eighty-eight percent of qualifying U.S. corn-producing counties experienced an increase in payment indemnity for losses due to excess wind, and nearly a third of those increases were statistically significant. Nearly the entire state of Nebraska has had a statistically significant increase in wind induced corn loss over the study period. Severe convective storms are the main cause of event-driven loss, as the peak climatology of significant hail and damaging convective wind gust events occurs during June–July across the Corn Belt (Gensini and Ashley 2011). Loss cost and acreage loss annual values have also experienced widespread increases across qualifying counties (Figs. 11e,f). Therefore, an important component going forward is to understand whether there has been, or is projected to be, an increase in the excess wind hazard itself, and also if corn producers need to consider this when obtaining crop insurance so as to avoid another 10 August 2020 scenario.
(iv) Cool-season perils
Cold wet weather indemnity for corn has undergone a robust increase in North and South Dakota with increases upward of $51,000 yr−1 (Fig. 11g). Elsewhere, there are scattered statistically significant increases throughout higher production areas of the Corn Belt and northern plains, including Nebraska, Iowa, and Illinois. Spring has especially experienced robust increases in precipitation in the northern plains and western domains of the Corn Belt and Lake regions (Easterling et al. 2017). Loss cost and acreage loss have both generally decreased in the northwestern and northern reaches of the study domain, whereas increases in indemnities were only subtle across the heart of the Corn Belt (Figs. 11h,i), which was consistent with previous recent literature.
As noted earlier, freeze/frost was the only peril to have a decrease in payment indemnity at the U.S. level, and it became clear that the decreases were widespread throughout the Corn Belt, albeit only a few reaching statistical significance (Fig. 11j). Some of the largest decreases in Nebraska reached nearly $17,000 yr−1. The western and northern reaches of the study domain, including western Nebraska, North Dakota, and northern Minnesota had subtle increases in indemnity (Figs. 11k,l). These have also been regions where the number of frost days have decreased the most across the United States along with the change in the date of the last spring frost (Easterling 2002).
4. Discussion
While the trends discussed above do have a physical risk component (i.e., changes in the frequency of perils), changes in land use and the expanding footprint of acreage can also explain a significant amount of the change. This “expanding bull’s-eye” of agricultural land use was also examined, partially motivated by earlier works examining the expanding bull’s-eye of the human-built environment to tornadoes, hurricanes, and floods (Ashley et al. 2014; Strader and Ashley 2015; Ferguson and Ashley 2017; Freeman and Ashley 2017; Strader et al. 2017, 2018). Previous literature has quantified that bigger corn-producing states suffered more from climate extremes than smaller production states; therefore, higher corn-producing areas with a larger harvest area coupled with higher seeding density may have resulted in more damage (Li et al. 2019). In other words, the average unadjusted payment indemnity for all weather and climate perils compiled together are the highest in counties and/or areas where corn production on average is greater than 12 million bushels (bu; 1 bu ≈ 0.035 m3) is located (Figs. 12a,b,e). This expansion is related to the growth of major and minor corn-producing areas that expose more fields to the aforementioned perils—and consequently—increase the risk for more insured corn losses. As corn production has increased by 8 times since the 1940s (National Agricultural Statistics Service 2021), the overall exposure to perils has played a significant role in corn payment indemnity. Since 1989, the most robust increases in corn production have been scattered throughout the Corn Belt, northern and southern plains, and delta regions (Fig. 12d). Meanwhile, the most robust increases in corn planting since 1989 were focused in North Dakota and South Dakota, as well as along the intersection of Colorado, Kansas, and Nebraska (Fig. 12c). These robust increases have also coincided with consistently high interannual corn yield variability in North and South Dakota (Kucharik and Ramankutty 2005), and where some of the most robust increases in payment indemnity have occurred (Fig. 12f). In summary, if corn acreage continues to expand, or if the density of corn crops increase, production will continue to increase and in turn will increase production revenue and yield; consequently, the commingling of these metrics will increase the overall exposure to the perils discussed and potentially increase the payment indemnity. These findings echo those presented for other hazards and similar applications (e.g., Ashley et al. 2014; Strader and Ashley 2015; Ferguson and Ashley 2017; Freeman and Ashley 2017; Strader et al. 2017, 2018).
Elements of the agricultural bull’s-eye effect including (a) corn production (bu), (b) major and minor corn production acreage based on average harvest, (c) trends in corn planting (acres yr−1), (d) trends in corn production (bu yr−1), (e) corn payment indemnity average of all compiled weather and climate perils (USD), and (f) trends in all compiled weather and perils indemnities for corn, where black-outlined counties represent statistical significance.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Elements of the agricultural bull’s-eye effect including (a) corn production (bu), (b) major and minor corn production acreage based on average harvest, (c) trends in corn planting (acres yr−1), (d) trends in corn production (bu yr−1), (e) corn payment indemnity average of all compiled weather and climate perils (USD), and (f) trends in all compiled weather and perils indemnities for corn, where black-outlined counties represent statistical significance.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
Elements of the agricultural bull’s-eye effect including (a) corn production (bu), (b) major and minor corn production acreage based on average harvest, (c) trends in corn planting (acres yr−1), (d) trends in corn production (bu yr−1), (e) corn payment indemnity average of all compiled weather and climate perils (USD), and (f) trends in all compiled weather and perils indemnities for corn, where black-outlined counties represent statistical significance.
Citation: Journal of Applied Meteorology and Climatology 61, 8; 10.1175/JAMC-D-21-0245.1
The agricultural expanding bull’s-eye effect cannot alone explain the regional and national increases in indemnity. Increases in corn prices and the increasing revenue in corn production over the 32-yr period both play a nontrivial role in U.S. payment indemnity increases (National Agricultural Statistics Service 2021). A statistically significant correlation coefficient between these two metrics and payment indemnity is 0.78 and 0.69, respectively. In addition, agricultural management such as drainage characteristics and irrigation act to enhance or offset the severity of losses. In the case of excess moisture, states such as North Dakota, South Dakota, and Minnesota experience amplified losses due to slower evaporation rates under a generally cooler climate and because of poorly drained soils in the region (Fig. 4g; Li et al. 2019). Though, it was the delta region that experienced the most robust increases in all three metrics for excess moisture loss in corn (Figs. 10g–i). This emphasizes that crop modeling can be a vital risk management strategy to help determine potential yield departure and need for hedging. Corn yield models [e.g., Agricultural Model Intercomparison and Improvement Project (AgMIP)] tend to model changes accurately when modeling temperature. When moderately wet and extremely wet conditions exist, most models cannot capture the observed nonlinear responses to yield reduction. Instead, they often indicated either a slight reduction in yield or increasingly higher yield under extreme wet conditions (Li et al. 2019). This is an area of future research that could benefit producers, especially those considering whether or not hedging might be cost-effective.
As mentioned, the greatest increasing trends in payment indemnity (>$150,000 yr−1) have occurred outside the Corn Belt region. These areas possessed a higher average loss cost and acreage loss value for all weather and climate perils. This is attributed to minor corn acreage locations often considered to be high risk in production, consequently, causing higher premium rates for coverage in these areas, which impacts indemnity (Coble and Barnett 2013; U.S. Government Accountability Office 2015). From 2005 to 2013, government costs per dollar of crop value were over 2.5 times as high in high-risk production areas as in low-risk production areas (U.S. Government Accountability Office 2015). As a result, loss cost and acreage loss values were generally higher in these areas than in the heart of the Corn Belt. Also, these were the regions where the most robust decreases in loss cost and acreage loss were located. As for trends, in addition to the potential increases in the peril intensity itself, the utilization of crop insurance by corn producers is a contributing factor for any trends as well. Over the study period, the number of indemnity payouts has increased by 2 times as a result of 150 million more acres being insured (RMA 2018; Coble and Barnett 2013), and thus the increase in participation will drive nonnormalized indemnity losses to increase. Previous literature has noted that a decreasing trend in loss cost could mean that hedging against a peril may be taking place due to perceived risk of higher occurrence of a certain peril in the region (Niles et al. 2019; Reyes and Elias 2019; Reyes et al. 2020). Furthermore, decreases in loss cost are due to either decreasing indemnities relative to liabilities, or increasing liabilities relative to indemnities (Reyes et al. 2020). Statistically significant decreasing trends in loss cost at the U.S. level were only noted for freeze/frost perils, and in this case, indemnity trends were decreasing. Subtle trends (statistically insignificant) were observed for drought, flood, and hail perils, which indicates hedging may be taking place for many counties, especially ones deemed as high risk with decreasing trends.
5. Conclusions
The use of normalization techniques, such as loss cost and acreage loss, on USDA Risk Management Agency data was effective in evaluating agricultural hazards at a relatively high spatiotemporal resolution. The Cause of Loss and Summary of Business datasets were additionally utilized to assess water deficit, water surplus, severe weather, and cool-season perils to further our understanding of crop insurance trends. Payment indemnity trends by peril are an indicator of both the biophysical and socioeconomic conditions surrounding corn losses in the United States. This study built upon previous literature to expand risk assessment related to weather and climate perils. One peril worth noting is excessive moisture. Given that the climatic trends in this hazard have substantially increased over the study period (1989–2020) and are projected to further increase, this peril is likely riskier than what may have been concluded previously. This is supported by statistically significant increases in payment indemnity, loss cost, and acreage loss for this peril, which were found to be the most robust of all perils examined. The delta region displayed the most robust of these spatial trends, which warrants further analyses into this regional risk. Corn producers could use the results of this study to analyze historic corn loss trends that may prompt various decisions (e.g., crop management, insurance participation), as results highlight potential adaption areas that would help increase resilience by peril by county. For example, integrating long-term trends in historic corn loss by weather and climate peril with near-term risk management practices is one practical application. Therefore, the examination of county-level aggregation of indemnities by peril over an annual basis provides further insight into the landscape of potential trends due to the changes in the peril frequency and/or socioeconomic conditions. Furthermore, county-level aggregation allows for identifying specific areas where corn insurance is being used to hedge against certain perils, where high indemnities or increasing loss cost and acreage loss trends are located, and where tactical and strategic decision-making needs to be made based on historical patterns of loss. As climate change continues to alter risk in the weather and climate peril landscape, producers need relevant information to better inform perception and knowledge of the implications these events bring to U.S. agriculture. The likelihood of increasing extreme weather and climate-driven events, coupled with the extreme events already witnessed over the last decade, is an indicator that federal crop insurance remains an important safety net for producers. Crop-loss analyses need to be part of a continuing effort to ensure food security for an increasing population.
Acknowledgments.
The authors thank the U.S. Department of Agriculture (USDA) Risk Management Agency (RMA) for making the data publicly available. Also, thanks are given to the anonymous reviewers whose comments and suggestions greatly enhanced the quality of this paper.
Data availability statement.
Data are open source and were obtained online (https://www.rma.usda.gov/Information-Tools/Summary-of-Business).
REFERENCES
Allen, J. T., M. K. Tippett, and A. H. Sobel, 2015: An empirical model relating us monthly hail occurrence to large-scale meteorological environment. J. Adv. Model. Earth Syst., 7, 226–243, https://doi.org/10.1002/2014MS000397.
Angel, J. R., and Coauthors, 2018: Midwest. Impacts, Risks, and Adaptation in the United States: Fourth National Climate Assessment, Vol. II, D. R. Reidmiller et al., Eds., U.S. Global Change Research Program, 872–940, https://doi.org/10.7930/NCA4.2018.CH21.
Ashley, W. S., and T. L. Mote, 2005: Derecho hazards in the United States. Bull. Amer. Meteor. Soc., 86, 1577–1592, https://doi.org/10.1175/BAMS-86-11-1577.
Ashley, W. S., S. Strader, T. Rosencrants, and A. J. Krmenec, 2014: Spatiotemporal changes in tornado hazard exposure: The case of the expanding bull’s-eye effect in Chicago, Illinois. Wea. Climate Soc., 6, 175–193, https://doi.org/10.1175/WCAS-D-13-00047.1.
Barthel, F., and E. Neumayer, 2012: A trend analysis of normalized insured damage from natural disasters. Climatic Change, 113, 215–237, https://doi.org/10.1007/s10584-011-0331-2.
Bengtsson, L., 2010: The global atmospheric water cycle. Environ. Res. Lett., 5, 025202, https://doi.org/10.1088/1748-9326/5/2/025202.
Berardi, D., M. D. Hartman, E. H. DeLucia, and T. W. Hudiburg, 2019: Flooding in the U.S. Corn Belt: Mitigating climate change and crop loss by converting to flood tolerant bioenergy crops. 2019 Fall Meeting, San Francisco, CA, Amer. Geophys. Union, Abstract B33E-04.
Bouwer, L. M., 2011: Have disaster losses increased due to anthropogenic climate change? Bull. Amer. Meteor. Soc., 92, 39–46, https://doi.org/10.1175/2010BAMS3092.1.
Brooks, H. E., and D. J. Stensrud, 2000: Climatology of heavy rain events in the United States from hourly precipitation observations. Mon. Wea. Rev., 128, 1194–1201, https://doi.org/10.1175/1520-0493(2000)128<1194:COHREI>2.0.CO;2.
Bruns, H. A., and H. Abbas, 2006: Planting date effects on Bt and non-Bt corn in the mid-south USA. Agron. J., 98, 100–106, https://doi.org/10.2134/agronj2005.0143.
Bundy, L. R., and V. A. Gensini, 2022: An assessment of USDA corn condition ratings across the U.S. Corn Belt. Agron. J., 114, 601–617, https://doi.org/10.1002/agj2.20973.
Cayan, D. R., T. Das, D. W. Pierce, T. P. Barnett, M. Tyree, and A. Gershunov, 2010: Future dryness in the southwest US and the hydrology of the early 21st century drought. Proc. Natl. Acad. Sci. USA, 107, 21 271–21 276, https://doi.org/10.1073/pnas.0912391107.
Changnon, D., and S. A. Changnon, 1997: Surrogate data to estimate crop-hail loss. J. Appl. Meteor., 36, 1202–1210, https://doi.org/10.1175/1520-0450(1997)036<1202:SDTECH>2.0.CO;2.
Changnon, D., and V. A. Gensini, 2019: Changing spatiotemporal patterns of 5-and 10-day Illinois heavy precipitation amounts, 1900–2018. J. Appl. Meteor. Climatol., 58, 1523–1533, https://doi.org/10.1175/JAMC-D-18-0335.1.
Changnon, S. A., Jr., 1971: Hailfall characteristics related to crop damage. J. Appl. Meteor., 10, 270–274, https://doi.org/10.1175/1520-0450(1971)010<0270:HCRTCD>2.0.CO;2.
Changnon, S. A., Jr., 2003: Present and future economic impacts of climate extremes in the United States. Global Environ. Change, 5B, 47–50, https://doi.org/10.1016/j.hazards.2004.04.001.
Changnon, S. A., and D. Changnon, 2000: Long-term fluctuations in hail incidences in the United States. J. Climate, 13, 658–664, https://doi.org/10.1175/1520-0442(2000)013<0658:LTFIHI>2.0.CO;2.
Changnon, S. A., and G. J. Hewings, 2001: Losses from weather extremes in the United States. Nat. Hazards Rev., 2, 113–123, https://doi.org/10.1061/(ASCE)1527-6988(2001)2:3(113).
Changnon, S. A., R. A. Pielke Jr., D. Changnon, R. T. Sylves, and R. Pulwarty, 2000: Human factors explain the increased losses from weather and climate extremes. Bull. Amer. Meteor. Soc., 81, 437–442, https://doi.org/10.1175/1520-0477(2000)081<0437:HFETIL>2.3.CO;2.
Changnon, S. D., 2003: Measures of economic impacts of weather extremes: Getting better but far from what is needed—A call for action. Bull. Amer. Meteor. Soc., 84, 1231–1236, https://doi.org/10.1175/BAMS-84-9-1231.
Cintineo, J. L., T. M. Smith, V. Lakshmanan, H. E. Brooks, and K. L. Ortega, 2012: An objective high-resolution hail climatology of the contiguous United States. Wea. Forecasting, 27, 1235–1248, https://doi.org/10.1175/WAF-D-11-00151.1.
Cleugh, H., J. Miller, and M. Böhm, 1998: Direct mechanical effects of wind on crops. Agrofor. Syst., 41, 85–112, https://doi.org/10.1023/A:1006067721039.
Coble, K. H., and B. J. Barnett, 2013: Why do we subsidize crop insurance? Amer. J. Agric. Econ., 95, 498–504, https://doi.org/10.1093/ajae/aas093.
Congressional Budget Office, 2017: Options to reduce the budgetary costs of the Federal Crop Insurance Program. CBO Rep. 53375, 38 pp., https://www.cbo.gov/system/files?file=115th-congress-2017-2018/reports/53375-federalcropinsuranceprogram.pdf.
Cook, B. I., T. R. Ault, and J. E. Smerdon, 2015: Unprecedented 21st century drought risk in the American southwest and central plains. Sci. Adv., 1, e1400082, https://doi.org/10.1126/sciadv.1400082.
Cooper, M., C. Gho, R. Leafgren, T. Tang, and C. Messina, 2014: Breeding drought-tolerant maize hybrids for the US corn-belt: Discovery to product. J. Exp. Bot., 65, 6191–6204, https://doi.org/10.1093/jxb/eru064.
Cooter, E., J. Bash, V. Benson, and L. Ran, 2012: Linking agricultural crop management and air quality models for regional to national-scale nitrogen assessments. Biogeosciences, 9, 4023–4035, https://doi.org/10.5194/bg-9-4023-2012.
Easterling, D. R., 2002: Recent changes in frost days and the frost-free season in the United States. Bull. Amer. Meteor. Soc., 83, 1327–1332, https://doi.org/10.1175/1520-0477-83.9.1327.
Easterling, D. R., and Coauthors, 2017: Precipitation change in the United States. Climate Science Special Report: Fourth National Climate Assessment, Vol. I, D. J. Wuebbles et al., Eds., U.S. Global Change Research Program, 207–230, https://doi.org/10.7930/J0H993CC.
Economic Research Service, 2020: Feedgrains sector at a glance. USDA, accessed February 2021, https://www.ers.usda.gov/topics/crops/corn-and-other-feedgrains/feedgrains-sector-at-a-glance/.
Elias, E., J. Reyes, C. Steele, and A. Rango, 2018: Diverse landscapes, diverse risks: Synthesis of the special issue on climate change and adaptive capacity in a hotter, drier southwestern United States. Climatic Change, 148, 339–353, https://doi.org/10.1007/s10584-018-2219-x.
Ferguson, A. P., and W. S. Ashley, 2017: Spatiotemporal analysis of residential flood exposure in the Atlanta, Georgia metropolitan area. Nat. Hazards, 87, 989–1016, https://doi.org/10.1007/s11069-017-2806-6.
Freeman, A. C., and W. S. Ashley, 2017: Changes in the US hurricane disaster landscape: The relationship between risk and exposure. Nat. Hazards, 88, 659–682, https://doi.org/10.1007/s11069-017-2885-4.
Ganguli, P., and A. R. Ganguly, 2016: Space-time trends in U.S. meteorological droughts. J. Hydrol. Reg. Stud., 8, 235–259, https://doi.org/10.1016/j.ejrh.2016.09.004.
Gensini, V. A., and W. S. Ashley, 2011: Climatology of potentially severe convective environments from the North American Regional Reanalysis. Electron. J. Severe Storms Meteor., 6 (8), https://ejssm.com/ojs/index.php/site/article/view/35/35.
Gensini, V. A., and H. E. Brooks, 2018: Spatial trends in United States tornado frequency. npj Climate Atmos. Sci., 1, 38, https://doi.org/doi:10.1038/s41612-018-0048-2.
Gensini, V. A., A. M. Haberlie, and P. T. Marsh, 2020: Practically perfect hindcasts of severe convective storms. Bull. Amer. Meteor. Soc., 101, E1259–E1278, https://doi.org/10.1175/BAMS-D-19-0321.1.
Glauber, J. W., 2013: The growth of the federal crop insurance program, 1990–2011. Amer. J. Agric. Econ., 95, 482–488, https://doi.org/10.1093/ajae/aas091.
Guastini, C. T., and L. F. Bosart, 2016: Analysis of a progressive derecho climatology and associated formation environments. Mon. Wea. Rev., 144, 1363–1382, https://doi.org/10.1175/MWR-D-15-0256.1.
Hallegatte, S., J.-C. Hourcade, and P. Dumas, 2007: Why economic dynamics matter in assessing climate change damages: Illustration on extreme events. Ecol. Econ., 62, 330–340, https://doi.org/10.1016/j.ecolecon.2006.06.006.
Herrero, M. P., and R. Johnson, 1980: High temperature stress and pollen viability of maize 1. Crop Sci., 20, 796–800, https://doi.org/10.2135/cropsci1980.0011183X002000060030x.
Höppe, P., and R. Pielke Jr., Eds., 2006: Workshop on climate change and disaster losses: Understanding and attributing trends and projections. GKSS Research Center Rep., 270 pp., https://sciencepolicy.colorado.edu/research_areas/sparc/research/projects/extreme_events/munich_workshop/full_workshop_report.pdf.
IPCC, 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
IPCC, 2012a: Summary for policymakers. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation, C. B. Field et al., Eds., Cambridge University Press, 1–19, https://www.ipcc.ch/site/assets/uploads/2018/03/SREX_FD_SPM_final-2.pdf.
IPCC, 2012b: Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. C. B. Field et al., Eds., Cambridge University Press, 582 pp., https://www.ipcc.ch/site/assets/uploads/2018/03/SREX_Full_Report-1.pdf.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp., https://doi.org/10.1017/CBO9781107415324.
IPCC, 2021: Climate Change 2021: The Physical Science Basis. Cambridge University Press, 2400 pp., https://www.ipcc.ch/report/ar6/wg1/.
Karl, T. R., and R. W. Knight, 1998: Secular trends of precipitation amount, frequency, and intensity in the United States. Bull. Amer. Meteor. Soc., 79, 231–242, https://doi.org/10.1175/1520-0477(1998)079<0231:STOPAF>2.0.CO;2.
Kistner, E., O. Kellner, J. Andresen, D. Todey, and L. W. Morton, 2018: Vulnerability of specialty crops to short-term climatic variability and adaptation strategies in the Midwestern USA. Climatic Change, 146, 145–158, https://doi.org/10.1007/s10584-017-2066-1.
Kucharik, C. J., and N. Ramankutty, 2005: Trends and variability in US Corn yields over the twentieth century. Earth Interact., 9, https://doi.org/10.1175/EI098.1.
Kunkel, K. E., 2003: North American trends in extreme precipitation. Nat. Hazards, 29, 291–305, https://doi.org/10.1023/A:1023694115864.
Kunkel, K. E., R. A. Pielke Jr., and S. A. Changnon, 1999: Temporal fluctuations in weather and climate extremes that cause economic and human health impacts: A review. Bull. Amer. Meteor. Soc., 80, 1077–1098, https://doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2.
Lauer, J., 2001: Earlier planting dates for corn: Real progress or an effect of global warming? Wis. Crop Manager, 8, 83–85.
Lesk, C., P. Rowhani, and N. Ramankutty, 2016: Influence of extreme weather disasters on global crop production. Nature, 529, 84–87, https://doi.org/10.1038/nature16467.
Li, Y., K. Guan, G. D. Schnitkey, E. DeLucia, and B. Peng, 2019: Excessive rainfall leads to maize yield loss of a comparable magnitude to extreme drought in the United States. Global Change Biol., 25, 2325–2337, https://doi.org/10.1111/gcb.14628.
Lobell, D. B., A. Torney, and C. B. Field, 2011: Climate extremes in California agriculture. Climatic Change, 109, 355–363, https://doi.org/10.1007/s10584-011-0304-5.
Lonnquist, J. H., and R. W. Jugenheimer, 1942: Factors affecting the success of pollination in corn. Agron. J., 35, 923–933, https://doi.org/10.2134/agronj1943.00021962003500110002x.
Mase, A. S., B. M. Gramig, and L. S. Prokopy, 2017: Climate change beliefs, risk perceptions, and adaptation behavior among Midwestern U.S. crop farmers. Climate Risk Manage., 15, 8–17, https://doi.org/10.1016/j.crm.2016.11.004.
Meehl, G. A., C. Tebaldi, G. Walton, D. Easterling, and L. McDaniel, 2009: Relative increase of record high maximum temperatures compared to record low minimum temperatures in the U.S. Geophys. Res. Lett., 36, L23701, https://doi.org/10.1029/2009GL040736.
Mieno, T., C. G. Walters, and L. E. Fulginiti, 2018: Input use under crop insurance: The role of actual production history. Amer. J. Agric. Econ., 100, 1469–1485, https://doi.org/10.1093/ajae/aay040.
Milly, P. C. D., R. T. Wetherald, K. Dunne, and T. L. Delworth, 2002: Increasing risk of great floods in a changing climate. Nature, 415, 514–517, https://doi.org/10.1038/415514a.
National Agricultural Statistics Service, 2019: Crop production. USDA, https://www.usda.gov/topics/farming/crop-production.
National Agricultural Statistics Service, 2021: National agricultural statistics service: Quick stats. USDA, https://www.nass.usda.gov/Quick_Stats/.
NCEI, 2021: Climate at a glance. NOAA, https://www.ncdc.noaa.gov/cag/.
Nicholls, N., 1996: Long-term climate monitoring and extreme events. Long-Term Climate Monitoring by the Global Climate Observing System, Springer, 101–115.
Niles, M. T., S. Wiener, R. E. Schattman, G. Roesch-McNally, and J. Reyes, 2019: Seeing is not always believing: Crop loss and climate change perceptions among farm advisors. Environ. Res. Lett., 14, 044003, https://doi.org/10.1088/1748-9326/aafbb6.
Perry, E. D., J. Yu, and J. Tack, 2020: Using insurance data to quantify the multidimensional impacts of warming temperatures on yield risk. Nat. Commun., 11, 4542, https://doi.org/10.1038/s41467-020-17707-2.
Pielke, R. A., 2005: Attribution of disaster losses. Science, 310, 1615–1616, https://doi.org/10.1126/science.310.5754.1615c.
Pryor, S. C., D. Scavia, C. Downer, M. Gaden, L. Iverson, R. Nordstrom, J. Patz, and G. P. Robertson, 2014: Midwest. Climate change impacts in the United States: The third national climate assessment. National Climate Assessment Report, J. M. Melillo, T. C. Richmond, and G. W. Yohe, Eds., US Global Change Research Program, 418–440.
Reyes, J. J., and E. Elias, 2019: Spatio-temporal variation of crop loss in the United States from 2001 to 2016. Environ. Res. Lett., 14, 074017, https://doi.org/10.1088/1748-9326/ab1ac9.
Reyes, J. J., E. Elias, E. Haacker, A. Kremen, L. Parker, and C. Rottler, 2020: Assessing agricultural risk management using historic crop insurance loss data over the Ogallala Aquifer. Agric. Water Manage., 232, 106000, https://doi.org/10.1016/j.agwat.2020.106000.
RMA, 2018: Loss adjustment manual standards handbook. USDA Doc., 419 pp., https://www.rma.usda.gov/-/media/RMA/Handbooks/Loss-Adjustment-Standards---25000/Loss-Adjustment-Manual/2018-25010-2H-Loss-Adjustment-Manual.ashx.
RMA, 2021: Cause of loss information—Summary of business data. USDA, accessed 15 March 2021, https://www.rma.usda.gov/SummaryOfBusiness/CauseOfLoss.
Rohli, R. V., N. Bushra, N. S. Lam, L. Zou, V. Mihunov, M. A. Reams, and J. E. Argote, 2016: Drought indices as drought predictors in the south-central USA. Nat. Hazards, 83, 1567–1582, https://doi.org/10.1007/s11069-016-2376-z.
Rosa, I., 2018: Federal Crop Insurance: Program overview for the 115th Congress. Congressional Research Service Rep. R45193, 30 pp., https://crsreports.congress.gov/product/pdf/R/R45193/4.
Rosenzweig, C., F. N. Tubiello, R. Goldberg, E. Mills, and J. Bloomfield, 2002: Increased crop damage in the US from excess precipitation under climate change. Global Environ. Change, 12, 197–202, https://doi.org/10.1016/S0959-3780(02)00008-0.
Scanlon, B. R., C. C. Faunt, L. Longuevergne, R. C. Reedy, W. M. Alley, V. L. McGuire, and P. B. McMahon, 2012: Groundwater depletion and sustainability of irrigation in the US High Plains and Central Valley. Proc. Natl. Acad. Sci. USA, 109, 9320–9325, https://doi.org/10.1073/pnas.1200311109.
Schaefer, J. T., J. J. Levit, S. J. Weiss, and D. W. McCarthy, 2004: The frequency of large hail over the contiguous United States. 14th Conf. on Applied Climatology, Seattle, WA, Amer. Meteor. Soc., 3.3, https://ams.confex.com/ams/pdfpapers/69834.pdf.
Schlenker, W., and M. J. Roberts, 2009: Nonlinear temperature effects indicate severe damages to US crop yields under climate change. Proc. Natl. Acad. Sci. USA, 106, 15 594–15 598, https://doi.org/10.1073/pnas.0906865106.
Shields, D. A., 2015: Federal crop insurance: Background. Congressional Research Service Rep. R40532, 23 pp., https://crsreports.congress.gov/product/pdf/R/R40532.
Smith, A. B., and R. W. Katz, 2013: US billion-dollar weather and climate disasters: Data sources, trends, accuracy and biases. Nat. Hazards, 67, 387–410, https://doi.org/10.1007/s11069-013-0566-5.
Smith, A. B., and J. L. Matthews, 2015: Quantifying uncertainty and variable sensitivity within the US billion-dollar weather and climate disaster cost estimates. Nat. Hazards, 77, 1829–1851, https://doi.org/10.1007/s11069-015-1678-x.
Steele, R., and J. L. Hatfield, 2018: Navigating climate-related challenges on working lands: A special issue by the USDA Climate Hubs and their partners. Climatic Change, 146, 1–3, https://doi.org/10.1007/s10584-017-2129-3.
Steiner, J. L., D. D. Briske, D. P. Brown, and C. M. Rottler, 2018: Vulnerability of southern plains agriculture to climate change. Climatic Change, 146, 201–218, https://doi.org/10.1007/s10584-017-1965-5.
Strader, S. M., and W. S. Ashley, 2015: The expanding bull’s-eye effect. Weatherwise, 68, 23–29, https://doi.org/10.1080/00431672.2015.1067108.
Strader, S. M., W. S. Ashley, T. J. Pingel, and A. J. Krmenec, 2017: Projected 21st century changes in tornado exposure, risk, and disaster potential. Climatic Change, 141, 301–313, https://doi.org/10.1007/s10584-017-1905-4.
Strader, S. M., W. S. Ashley, T. J. Pingel, and A. J. Krmenec, 2018: How land use alters the tornado disaster landscape. Appl. Geogr., 94, 18–29, https://doi.org/10.1016/j.apgeog.2018.03.005.
Tang, B. H., V. A. Gensini, and C. R. Homeyer, 2019: Trends in United States large hail environments and observations. npj Climate Atmos. Sci., 2, 45, http://doi.org/10.1038/s41612-019-0103-7.
Thornton, P. K., P. J. Ericksen, M. Herrero, and A. J. Challinor, 2014: Climate variability and vulnerability to climate change: A review. Global Change Biol., 20, 3313–3328, https://doi.org/10.1111/gcb.12581.
United Nations, 2019: Population facts. Department of Economic and Social Affairs Doc., 4 pp., https://www.un.org/en/development/desa/population/publications/pdf/popfacts/zPopFacts_2019-6.pdf.
U.S. Government Accountability Office, 2015: Crop insurance: In areas with higher crop production risks, costs are greater, and premiums may not cover expected losses. GAO Rep. GAO-15-215, 54 pp., https://www.gao.gov/assets/gao-15-215.pdf.
Vose, R. S., D. R. Easterling, and B. Gleason, 2005: Maximum and minimum temperature trends for the globe: An update through 2004. Geophys. Res. Lett., 32, L23822, https://doi.org/10.1029/2005GL024379.
Wallander, S., M. Aillery, D. Hellerstein, and M. Hand, 2013: The role of conservation programs in drought risk adaptation. USDA Economic Research Service Rep. ERR-148, 68 pp.
Walthall, C. L., and Coauthors, 2013: Climate change and agriculture in the United States: Effects and adaptation. USDA Tech. Bull. 1935, 186 pp., https://www.usda.gov/sites/default/files/documents/CC%20and%20Agriculture%20Report%20(02-04-2013)b.pdf.
Westcott, P. C., and M. Jewison, 2013: Weather effects on expected corn and soybean yields. USDA Economic Research Service Doc. FDS-13g-01, 31 pp., https://www.ers.usda.gov/webdocs/outlooks/36651/39297_fds-13g-01.pdf?v=737.
Wilcox, R. R., 2010: Fundamentals of Modern Statistical Methods: Substantially Improving Power and Accuracy. Springer, 249 pp.
