a. Overview of findings

Through our analysis of this literature published over the past 30 years, we were able to answer our research questions about agricultural decision makers’ perceptions and use of climate and weather information, forecasts, and DSTs. Variables and themes emerging from this literature included measured and hypothesized barriers to optimal application of weather/climate information and tools to agriculture, farmers’ and advisors’ perspectives on useful information, adaptations that farmers could make by using weather and climate forecasts, as well as aspects of the research itself—such as whether a participatory approach to forecasts/DST development was used or advocated. As shown below in Table 2, most studies focused on if and how farmers use climate information and/or forecasts and discussed barriers that are reducing farmers’ use of forecasts/DSTs in their decision making. Although not all of the articles measured use, 28 studies measured and/or discussed farmers’ perceptions of climate information/forecasts/DSTs.

Table 2.

Main themes and variables in this literature.

View Table

Advisors to farmers are less studied than farmers themselves, with only 12 articles addressing advisors’ perceptions, or use of climate information and tools, and factors limiting their utilization. Types of agricultural advisors studied included university extension, private crop consultants, and agribusinesses, such as seed dealers. Most of the included articles also discussed farmer and/or researcher perspectives on potential applications for climate information in agriculture and possible climate change adaptation options that could be informed by climate forecasts or tools.

More than half of the articles (28) emphasized the challenges of communicating with farmers and/or advisors about weather and climate information, and/or offered recommendations about how to improve communication on these topics. About 40% of the articles (18) relied upon participatory research or advocated participatory development of weather or climate tools to increase the likelihood that farmers will actually use the information to inform decisions. Benefits of participatory processes for weather and climate tool development were discussed in this literature, especially in those published in the last decade. Twelve studies highlighted farmer preferences for, as well as potential benefits and constraints of, using the Internet as a medium for communication of weather and climate information. Frisvold and Murugesan (2013) made the important point that while wealthier farmers with larger operations tend to have Internet access and use online resources in managing their farm, this is not necessarily the norm for smaller-scale and lower-income farmers.

b. Social–psychological factors affecting farmers’ perceptions and use of climate information

This literature brings to light a variety of social–psychological factors that play an important role in the incorporation of weather and climate forecasts into agriculture. One challenge for developers of forecasts and DSTs is helping farmers cope with uncertainty and prepare for variability by providing information that they will actually use. An issue established in the literature is the disconnect between timing of available weather and climate information and when farmers are making decisions (Ash et al. 2007; Changnon 2004; Furman et al. 2011; Keogh et al. 2005; McNew et al. 1991; Sonka et al. 1982, 1992). If forecasts about a given growing season are not available or able to accurately predict conditions until after a farmer has made a key decision, they cannot be effectively utilized and incorporated in decision making. Another longstanding factor highlighted in this literature is how farmers deal with uncertainty and weather variability when managing their business/farm, and also variation in farmers’ ability to interpret probabilities and uncertainty as a part of weather and climate forecasts (Breuer et al. 2008, 2009; Cabrera et al. 2008; Changnon et al. 1988; Crane et al. 2010; Everingham et al. 2002; Garbrecht and Schneider 2007; Getz 1978; Hansen 2002; Jagtap et al. 2002; Keogh et al. 2004, 2005; McCrea et al. 2005; Meinke and Stone 2005; Power et al. 2007; Roncoli 2006).

Farmers’ perceptions of the risk to their farms from weather/climate variability is another issue affecting how they consider and use climate and weather information. Hansen (2002) points out that “The primary motivation for individual farmers is an awareness of some level of vulnerability to impacts of climate variability, and opportunity to reduce that vulnerability through appropriate use of forecasts information” (p. 312). Fifteen articles included some discussion of how farmers manage or minimize risks to their operations from weather or climate variability (see Table 2). Crane et al. (2010) found that farmers know they cannot avoid all weather-related risks to their farm operation, and thus “employ management strategies ensuring some yield during most years and under most conditions…. The rationale for this approach is that consistency eventually pays off and… it is safer than trying to adjust cropping patterns seasonally to maximize short-term gain” (p. 50). As shown in Table 2, 10 articles discussed weather risk events (droughts, hurricanes, ice storms, and other extreme weather events that were experienced over the short term) that focused attention and influenced attitudes toward climate change/variability and risk management behaviors. Several articles explored how farmers view weather risks in economic terms, such as potential financial losses due to more variable precipitation, as well as the potential economic benefits of improved forecasts—particularly providing a competitive advantage in a global marketplace (Crane et al. 2010; Frisvold and Murugesan 2013; Jagtap et al. 2002; Marshall et al. 2011; McCrea et al. 2005; Meinke and Stone 2005; Nelson et al. 2002; Reid 2009; Tarnoczi and Berkes 2010).

A few articles (Furman et al. 2011; Marshall et al. 2011; Reid et al. 2007) highlighted the importance of social capital, arguing that when farmers have limited social capital it is more difficult for them to learn about new forecasting technologies and tools through informal interactions with other farmers. Crane et al. (2010) found that social networks were crucial for the dissemination and processing of information by farmers, often through social gatherings, more formalized farmer meetings, as well as business interactions with agricultural suppliers and brokers. Similarly, Artikov et al. (2006) and Hu et al. (2006) examined the role of perceived social norms in the decision to adopt new technologies or allow weather forecasts to influence decision making. Through multiple regression analysis, Artikov et al. (2006) and Hu et al. (2006) found that social norms had a significant influence on farmers’ use of weather information to make decisions, especially agronomic decisions. Another motivator for farmers discussed in the literature was being able to observe new practices implemented by other farmers and a resulting decrease in perceived financial risk (Tarnoczi and Berkes 2010; Marshall et al. 2011). Tarnoczi and Berkes (2010) conducted interviews with Canadian farmers about climate change adaptation and found support for the “diffusion of innovations” framework (Rogers 2003) since “Information that was observable or experiential was more significant for the adoption of new practices. For example, farmers claimed to be more willing to try conservation tillage techniques after seeing a neighbour succeed with the practice, even when they already had prior information on it” (p. 303).

Trust—in the credibility, accuracy, and source of information as well as in advisors or consultants—influences farmers’ perceptions and use of climate and weather forecasts and tools. The concepts of trust and source credibility are important in communication, especially when the source intends to persuade the audience to change their attitudes and/or behaviors, such as convincing farmers to use weather and climate information to inform decisions (Renn and Levine 1990). Hu et al. (2006) point out that farmers were skeptical of forecasts that contradicted one another, especially when it was unclear who or what organization had produced each forecast. Hansen et al. (2004) found an almost even split in confidence or lack of confidence in forecasts among Florida farmers. According to Marshall et al. (2011), most rangeland ranchers heard predictions of drought frequently, which lead them to be skeptical of when to take a prediction seriously after experiencing “false warnings.” The importance of trusted advisors or consultants who could help DST developers reach farmers and work with farmers to interpret climate information were addressed in a few of the articles. Trusted sources and potential outreach partners included family or business partners (Keogh et al. 2004), other farmers (Furman et al. 2011), Extension agents (Breuer et al. 2008, 2009; Cabrera et al. 2008; Dinon et al. 2012; Jagtap et al. 2002), climate scientists, and private agronomists and consultants (Carberry et al. 2002).

Collaborative approaches to developing forecasts and DSTs and participatory research stand out as an important strategy in this literature. As noted above in Table 2, more than a third of the included articles either utilized a participatory approach to tool development or advocated for more participatory processes and research endeavors. Bartels et al. (2013) describe in more detail the structure of the SECC participatory process that fostered two-way interactions between farmers, extension agents, and university researchers around climate change adaptation strategies. Asking for farmers’ input early in the process and seeking feedback throughout can allow farmers to feel a sense of ownership over tools and have more favorable attitudes toward adaptation options recommended by forecasts or DSTs. This “buy in” or sense of ownership can improve farmers’ and advisors’ perceptions of climate information and tools and is a promising strategy to increase their use (Breuer et al. 2009; Cabrera et al. 2008; Hansen 2002; Jagtap et al. 2002; Meinke and Stone 2005; Nelson et al. 2002; Reid 2009). Overall, this literature shows the promise of social science for understanding social psychological determinants of favorable attitudes and use of weather and climate information and forecasts.

c. Useful information for agricultural decision making

Several of the articles in this review asked farmers and/or their advisors about weather or climate information that has been or would be useful for agricultural decision making. Hartmann et al. (2002) point out that the most useful aspect of a forecast is likely to vary with the type of decision a farmer is making. For example, farmers in Georgia emphasized the importance of shorter-term weather information (such as for the next week), which they already use and may check on a daily basis during the growing season, relative to longer-term forecasts (monthly or seasonal outlooks) that were perceived to have lower accuracy (Crane et al. 2010). Rangeland managers in the southwestern United States are a promising audience for decision support tools, since they are already using historical precipitation information (73%) and seasonal precipitation forecasts (69%) to make decisions (Crimmins et al. 2007). Changnon (2004) reviewed his previous research on U.S. farmers and advisors including agricultural chemical manufacturers, finance companies, pest management consultants, etc., that measured their use of forecasts and found that “usage of climate predictions for general background information grew from 1981 to 1999, as did usage for making specific decisions” (p. 611). Similarly, Dinon et al. (2012) found that North Carolina Extension agents agree that predictions about the weather for the upcoming growing season would be helpful.

Regarding what information agricultural decision makers would like access to, Crimmins et al. (2007) found that southwestern U.S. “land managers’ needs for climate information [go] beyond traditional statistical averages. Needs appear to include information on potential changes to the likelihood, frequency, and duration of climate and weather extremes, important in long-term planning to reduce vulnerability to climate change” (p. 85). Similarly, some research revealed that farmers wanted forecasts that could accurately predict extreme weather events (Easterling 1986; Sonka et al. 1992). In Australia, agricultural consultants preferred applied “what does this mean?” information presented in combination with crop models rather than weather conditions or forecasts by themselves. In an interview, one Australian agricultural consultant said, “There is no point knowing how much water you’ve got in your profile if you don’t know what it means to your risk profile, or your cropping potential, if you like. Knowing what that level is is one thing, but knowing what it means is another thing…. That’s where crop models come in” (McCown et al. 2012, p. 37). Similarly, Hansen et al. (2004) found that Florida farmers wanted more applied information about how El Niño and La Niña would impact temperatures, especially winter lows. Together, these findings indicate that there is significant room to better combine and package weather and climate information and forecasts to suit the needs of agriculturalists.

Not surprisingly, several studies found that farmers and/or advisors would like to see improved forecast skill or reliability. Crane et al. (2010) found that while farmers were more likely to rely on short-term forecasts (such as 5-day weather forecasts) than seasonal climate forecasts, they did not completely trust this information to always be relevant and accurate. Many studies concluded that farmers want contextualized, location-specific forecast information to facilitate decision making on their own farms (Artikov et al. 2006; Breuer et al. 2008; Carberry et al. 2002; Carlson 1989; Changnon 1988, 2004; Crane et al. 2010; Furman et al. 2011; Hansen 2002; Hu et al. 2006; Keogh et al. 2005; McNew et al. 1991; Nelson et al. 2002; PytlikZillig et al. 2010; Roncoli 2006; Weiss and Robb 1986). While certain parameters, such as precipitation, may be difficult to predict at very local scales, forecasts and information can still be tailored to be more relevant to the context of a particular agricultural sector (i.e., frost forecasts for Florida fruit growers) (Westra and Ashish 2010). Additionally, they would like access to forecasts at the correct time to facilitate decisions (such as predicted harvest conditions) and predict impacts, so they can be mitigated before they become severe (Changnon 2004; Hansen 2002). While some of the characteristics that farmers want may be beyond the scope or ability of current forecasts, these findings point out areas where forecasts and tools can be improved and strive to achieve in the future. This literature also highlights current gaps between producers’ desires and what modelers and forecasters have provided thus far.

d. Agricultural adaptations that could be informed by climate forecasts/tools

Most of the articles in this review presented scientists’, advisors’, and/or farmers’ opinions about how climate forecasts or DSTs could be used to aid climate adaptation decision making. Potential uses for agronomic decisions ranged from changing crops and crop varieties (Crane et al. 2010), determining stocking rates for rangeland livestock (Keogh et al. 2004, 2005), choosing whether or not to plant winter forage for cattle (Jagtap et al. 2002), and deciding if/when to plough sugar cane (Everingham et al. 2002) to Australian wheat farmers using the Southern Oscillation index (SOI) as an indication of increased rainfall and deciding to plant early to benefit from improved growing conditions (Nelson et al. 2002). Ash et al. (2007) argue that utilizing forecasts proactively to take advantage of favorable conditions may be more important for agriculture than using them to prepare for poor growing conditions. Furman et al. (2011) described how organic farmers, often with more flexible operations than conventional farmers, make initial plans around available 10-day forecasts, and then when short-term forecasts and additional data were available they could modify their management activities and decisions if needed. Several articles touched on the potential for farmers to reduce economic risks or increase their competitive advantage in increasingly international markets by using climate or weather forecasts. They also noted farmers’ interest in weather conditions in other regions of the world that are producing the same crops and how conditions will affect commodity prices (Changnon 2004; Crane et al. 2010; Hansen 2002; Meinke and Stone 2005). Hartmann et al. (2002) heard informally from discussions with farmers that they are interested in how competitors might use forecasts, but remained skeptical of their value owing to perceived low accuracy. Overall, the research conducted thus far points to several avenues for improved weather and climate forecasts and DSTs to aid farmers in adapting to climate change.

e. Barriers limiting farmers’ use of climate information

Many of the articles emphasized underutilization of weather and climate forecasts in agriculture relative to their potential (Ash et al. 2007; McCown et al. 2012). For example, Marshall et al. (2011) point out that, although Australian rangeland cattle ranchers “showed interest in the economic and environmental benefits and costs of seasonal climate forecasts, only 40% of graziers in our study were highly likely to adopt the [SCF] technology once it becomes available to the region” (p. 522). More than half of the articles discussed barriers to optimal use of climate information or tools by farmers. Some studies asked farmers or their advisors directly, while others drew inferences from other data about why climate information, forecasts, and tools are underutilized by farmers in the developed world. Barriers to use fell into three broad categories: aspects of the forecast itself (such as skill, scale, and lead time), farmers and advisor perceptions (such as interpretation/understanding, risk from weather/climate, and urban bias), and contextual factors (such as inflexible operations, market fluctuations, and lack of resources).

The most commonly cited barrier, especially in more recent research, was farmers’ or advisors’ perceptions of low accuracy of forecasts/models and their concern that predictions and information were not provided on a local enough scale to be useful (Artikov et al. 2006; Breuer et al. 2008; Cabrera et al. 2008; Carberry et al. 2002; Changnon et al. 1988; Crane et al. 2010; Dinon et al. 2012; Easterling 1986; Hansen et al. 2004; Hartmann et al. 2002; Hu et al. 2006; Keogh et al. 2005; Marshall et al. 2011; McCrea et al. 2005; Power et al. 2007; Sonka et al. 1992; Vining et al. 1984; Weiss and Robb 1986). Farmers’ lack of understanding or awareness of available climate information was addressed in 10 articles (Cabrera et al. 2006; Changnon et al. 1988; Crane et al. 2010; Furman et al. 2011; Getz 1978; Keogh et al. 2004, 2005; Reid et al. 2007; Sonka et al. 1992; Vining et al. 1984). Through interviews in Florida, Hansen et al. (2004) found that farmers were not aware of their misconceptions in terms of weather and climate information and forecasts; in other words, they “don’t know what they don’t know,” and this overconfidence may reduce their motivation to understand forecasts more thoroughly.

Seven studies from the late 1970s through early 1990s found that the types of weather information that farmers indicated would be useful were unavailable in their main sources of weather information—public television and radio forecast—and that short forecast lead times impeded use (Carlson 1989; Changnon et al. 1988; Easterling 1986; GAO 1979; Getz 1978; McNew et al. 1991; Sonka et al. 1992). Easterling (1986) concluded that the lack of climate prediction lead time was the most important barrier to broader use of the National Oceanic and Atmospheric Administration (NOAA)’s monthly and seasonal weather outlook. A more recent study by Keogh et al. (2005) discussed how very short forecast lead times made it difficult for rangeland graziers to utilize weather information in decision making. These authors cite research from Africa conducted by Ingram et al. (2002), which suggested that “a less skillful forecast issued with sufficient lead time may be more valuable than a highly skillful forecasts that arrives at a time when a producer’s decision cannot be revoked” (Keogh et al. 2005, p. 1614). The specific preferred lead time varied depending on the study and within studies between different types of agricultural production and different decisions that had potential to be informed by weather and/or climate information, but was generally on the scale of a month to a few months. On a similar note, Crane et al. (2010) reported that farmers perceived an “urban bias” in forecasting, meaning that television stations tailor their weather information to cities rather than rural areas where agriculture occurs. A report by the GAO (1979) found that farmers preferred forecasts tailored to agriculture rather than forecasts designed for the general public.

Eight articles cited fixed or inflexible management, operations, or resource conditions as constraints on farmers’ use of climate information and forecasts (Breuer et al. 2008; Cabrera et al. 2006; Crane et al. 2010; Furman et al. 2011; Hansen 2002; Hu et al. 2006; Jagtap et al. 2002; Marshall et al. 2011). Greater concern with issues other than weather or climate risks, such as government regulation and market fluctuations, was noted in seven articles (Artikov et al. 2006; Breuer et al. 2008; Cabrera et al. 2008; Crane et al. 2010; Jagtap et al. 2002; Tarnoczi and Berkes 2010; Weiss and Robb 1986). Five articles discussed limitations posed by tools or forecasts that are too difficult or complicated to use, or require the help of an expert to interpret (Carberry et al. 2002; Garbrecht and Schneider 2007; Keogh et al. 2004, 2005; Power et al. 2007). Five of the included articles discussed that farmers or advisors did not recognize a problem that could be addressed with climate/weather information or tools (Cabrera et al. 2006, 2008; Carberry et al. 2002; Dinon et al. 2012; Reid et al. 2007) and five mentioned farmers’ aversion to new technology or change (Breuer et al. 2008; Cabrera et al. 2006; Changnon et al. 1988; Marshall et al. 2011; Reid et al. 2007). A related barrier was that some farmers remained unconvinced that changing decisions based on climate information would lead to improved agricultural outcomes (Changnon 2004; Marshall et al. 2011; Power et al. 2007; Roncoli 2006). In at least one study, farmers believed that others would benefit more from weather information: “Several farmers stated that other stakeholders, such as chemical suppliers, insurance companies, brokers, relief agencies, and banks, might benefit more from climate forecasts than would agricultural producers” (Breuer et al. 2008, p. 394).

A few articles identified government financial assistance, through crop insurance and other programs, as reducing incentives to adapt to climate variability and at times limiting farmers’ flexibility through strict requirements (Cabrera et al. 2007; Crane et al. 2010; Jagtap et al. 2002). An additional barrier is that farmers are making decisions in the short term, rather than the long term, and short-term decisions do not always require or benefit from seasonal climate forecasts or longer-term planning and adaptation tools (Cabrera et al. 2006; Roncoli 2006). Several studies found that farmers preferred and relied on short-term forecasts, such as 24 h or 5 day, rather than longer-term forecasts, suggesting that farmers are engaging in more short-term decision making (Breuer et al. 2008; Crane et al. 2010; Hansen et al. 2004; McNew et al. 1991; Weiss and Robb 1986). Two articles mentioned the high cost of recommended adaptations as a potential disincentive for utilizing weather forecasts and tools for adaptation planning (Breuer et al. 2008; Cabrera et al. 2006). A related barrier is that farmers planning to retire soon and not expecting children or other family to continue the operation are not motivated to make investments in new technologies or long-term risk management strategies, including those that would improve their farms’ adaptability to climate change (Reid et al. 2007).

There is an extensive literature on the potential of climate forecasts/DSTs for agriculture in developing countries, which we felt was too large and diverse to synthesize in this review of developed countries. Some general areas of overlap that could be further explored include preferences for contextualized and localized weather/climate information (Ash et al. 2007; Roncoli 2006) and the importance of timing of accurate forecasts before farmers make decisions, which can be one to six months in advance (Luseno et al. 2003; McIntosh et al. 2005). The importance of source credibility for climate information was highlighted for agricultural decision makers in developing countries as well as developed (Ingram et al. 2002). Another concept that was relevant in both the developing and developed world was the potential and importance of participatory approaches to weather/climate tool development (Gadgil et al. 2002; Patt et al. 2005; Podesta et al. 2002).

f. Strengths of existing research

Overall, this literature provides good coverage of seasonal climate forecast use, particularly in the southeastern United States and Australia. Collectively, these studies have explored a variety of issues with SCF use in a variety of agricultural contexts in these regions. Participatory decision support tool development processes have also been the subject of thorough study that describes the strengths and challenges of conducting them in each of these regions, and a case is made for the continued use of participatory processes to increase the use and effectiveness of climate DSTs in agriculture.

Our analysis leads us to conclude that involving farmers, researchers, advisors, and other agricultural stakeholders in the development of DSTs can have numerous benefits. Benefits of participatory processes highlighted in this literature include providing each stakeholder group with a more holistic view of the agricultural system, allowing participants to recognize common ground for improved communication, and improved adaptation to climate change as farmers face additional environmental regulations and economic hurdles (Cabrera et al. 2008). As Breuer et al. (2009) point out, the participatory process around the development of AgroClimate, a decision support system (DSS) for the southeastern United States, has led to higher awareness and broader distribution, and “Overall, participation enhances legitimacy of the project” (p. 9). Roncoli (2006) and Cabrera et al. (2008) caution that participatory research approaches have the potential to “introduce biases” into a product if they are not conducted with a representative population and could lead to tools that meet the needs of one group or subgroup more than other stakeholders, so we agree they should be used alongside additional “comparative and complementary methodologies to ensure consistency and integrity in the whole process” (Cabrera et al. 2008, p. 405).

One distinction worth noting is that some of the studies in this review measured actual use of climate information or DSTs by farmers and advisors, while other studies examined farmers’ or advisors’ perspectives on potential use/usefulness in the future. This distinction between research studies has implications for how results from these different studies can be interpreted. Understanding current use is most helpful for determining which available information and tools are working and which are not for farmers’ present needs. However, there is also value in understanding what the agricultural community perceives as potential uses for weather and climate information in the future. As long as these potential uses for weather or climate information are proposed by intended end users, there is a role for this information in future research and DST development. Future studies should continue to solicit this type of feedback from farmers and advisors in order to highlight applications and uses that forecasters and tool developers may not have considered.

g. Evolution of and gaps in this literature: Improving future research

Through completing this review, it became clear that the literature around use of climate and weather information in agriculture has evolved over time. Early studies focused on the potential of forecasts and problems with farmers’ awareness and ability to interpret available forecasts and information. A few articles emphasized the need to educate farmers or advisors to increase use of existing forecasts in their operations. Later studies highlight the importance of a more participatory approach, in which two-way communication is valued for developing and tailoring weather information/DSTs to users’ needs. This parallels findings from other areas of research on the importance of user involvement in the ultimate usability of scientific information, such as that reported in Dilling and Lemos (2011). Participatory approaches could be expanded to additional regions beyond the southeastern United States and Australia to cover more agricultural decisions and issues (Carberry et al. 2002). We agree with Cabrera et al. (2008), Roncoli (2006), and others who emphasized the need for interdisciplinary research teams to tackle this topic: “Participatory modeling for climate applications holds a great promise for future work but requires interdisciplinary teams to embrace the greater complexity among human, natural resource management, and climate systems” (Cabrera et al. 2008, p. 406).

From a methodological standpoint, early research tended to rely more exclusively on survey methodology, and appears to be less rigorous in terms of sample sizes and generalizability than later research. Research from the early 2000s on has shifted toward using mixed methods approaches, combining both quantitative (such as surveys) and qualitative (such as interviews and focus groups) research methods. This is a positive development that likely signals more thoughtful application of social science to these weather/climate information issues, since surveys are not a “one size fits all” methodology to answer every research question (Prokopy 2011). Future research should continue to critically apply the most appropriate methods to explore new research questions about farmers’ use and perceptions of weather and climate information and DSTs. For example, if researchers are interested in how broadly a particular DST has been used, a survey may be the best method, but if they want to know why a certain group of farmers are/are not using the DST, qualitative methods such as interviews would be more appropriate [for more on applying social science methods to DST development, see Prokopy et al. (2013b)].

This review revealed some weaknesses and gaps in this body of research. While there has been significant study of the southeastern United States, less research has explored farmers in other regions of the United States in terms of their perceptions and use of weather and climate DSTs. Similarly, little to no social science research has been conducted in Europe with farmers or agricultural advisors to explore their use of climate information and tools for decision making, particularly in response to global climate change. Since different regions and agricultural systems are subject to different climate projections and impacts, it is critical to understand the needs and behaviors of the diversity of U.S. and European farmers regarding weather and climate information and adaptation to climate change. A similar area requiring further study is how farmers operating in different types of agricultural systems in varied contexts may have different perceptions and uses of climate information or DSTs (Breuer et al. 2009; McNew et al. 1991). For example, large-scale corn and soybean farmers in the Midwest are likely to have different concerns and needs for weather and climate decision support information than organic vegetable producers in California. Further research should be conducted to understand which forecasts or tools different types of farmers are already using, when they need them, and for which decisions they are already using climate information (Hu et al. 2006) potentially as case studies exploring these issues with subgroups that are underserved or unconventional. Related areas requiring further study are the power and politics of weather and climate information and forecasts, in terms of who has access to the information and forecasts, and the role of power in who benefits most from these advancements. These topics have been explored more in developing countries (see Lemos et al. 2002; Broad and Agrawala 2000; Nelson and Finan 2009) but deserve to be considered in more industrialized contexts.

The focus in this literature on farmers more than advisors may be appropriate, but has left a gap. Although farmers are the ultimate decision makers in agriculture, advisors may be the “primary” users of climate information and tools as they can use this information and/or forecasts to enhance the advice they give to farmers. Farmers may not have the time, energy, interest, or capacity to convert weather forecasts and tools into usable information for decision making. Advisors have the potential to act as brokers of this type of information as they guide farmers toward decisions. Future studies should explore more types of agricultural advisors and their perspectives and uses for weather/climate DSTs.

Much of the existing literature examines farmers’ perceptions and use of seasonal climate forecasts specifically, while less research has addressed more general weather or climate information and DSTs. To address this gap, future research should focus on other types of weather forecasts and information that may be more relevant to particular issues or contexts. For example, SCFs based on El Niño are more effective and relevant in some regions such as the southeastern United States, but broadening the scope of types of forecasts and information examined in research could identify better ways to address the needs of agricultural decision makers across the United States and other developed nations. Research conducted in additional agricultural contexts with new variables drawing from more models and forecasts would be valuable (Jochec et al. 2001; McNew et al. 1991).

This review highlighted how social science concepts and theories such as social capital, social networks, social norms, and the diffusion of innovations have guided researchers’ understanding of perceptions, use, and barriers to broader use of climate information and DSTs in agriculture. These theories and others can help research in this area go further in the future, since research solidly grounded in social theory is somewhat sparse. Several authors—including Furman et al. (2011), Hu et al. (2006), Jagtap et al. (2002), and PytlikZillig et al. (2010)—saw a need to better understand the social relationships that influence farmer decision making as well as the role of trust in farmers’ use of climate information. There is also a gap in the literature in terms of farmers’ attitudes toward the usefulness of climate forecasts, how they form, and the role of past experiences on attitudes toward climate forecast (McCrea et al. 2005). Future studies could more explicitly use theoretical frameworks, such as theories of risk perception, the diffusion of innovations framework, or theory of planned behavior/reasoned action approach to guide research on these topics.