1 Introduction

Drought can have significant adverse effects on a range of socio-economic and natural systems, with an estimated 55 million people globally affected by drought every year (WHO 2021). Drought is a complex and contextually varying natural hazard that is broadly defined as a prolonged absence of precipitation (IPCC, 2007; Singh 2024). Due to its complexity, accurately identifying and adapting to drought impacts is a challenging, but critical, task for drought-affected communities. Typically, drought has been classified into four main categories: meteorological, agricultural, hydrological, and socio-economic (Hayes et al. 2007; Xiao et al. 2024). Though all types of droughts have severe impacts on society, agricultural drought is a particular hazard to Australia’s population and economy. Agricultural drought refers to conditions when soil moisture is insufficient for crop growth and production (Sardar et al. 2020). While drought is a ubiquitous part of the Australian environment, climate projections indicate that the frequency and severity of drought is likely to increase due to anthropogenic climate change (BoM and CSIRO, 2020; IPCC 2014). Thus, drought will have significant impacts on vulnerable regions prone to land degradation, leading to heighted pressure on agricultural systems and communities (Andersson et al. 2020).

The Murray-Darling Basin (MDB) is Australia’s most valuable agricultural region contributing to almost 40% of national agricultural production (Abel et al. 2016). The MDB has experienced severe droughts in the past, including the Millennium Drought (1997–2009) and recently the 2017–2019 drought (BoM 2020). Drought will continue to have significant negative impacts on natural resources, agricultural productivity, and economic and social activities in the MDB (Kiem 2013; Berbel and Esteban 2019). In a drought-vulnerable future, there is growing pressure on institutional bodies of all levels to implement adaptation and mitigation measures for future drought impacts.

Historically, drought has been managed in a reactive manner. Such management approaches have been criticized in the literature, and proactive drought management strategies are now emerging (Nhamo et al. 2019). To minimize drought impacts, approaches are shifting from humanitarian crisis-driven strategies toward preparedness and community resilience-building as part of disaster risk reduction (Calvel et al. 2020). Studies have shown that drought or disaster relief does little to reduce societal vulnerability to the next drought event, demonstrating the need for preventive measures (Wilhite et al. 2014). Vulnerability to drought is often the direct result of poor planning and resource management practices. To reduce societal vulnerability, it is important to promote proactive action, planning, and resource management by redirecting resources from response programs to mitigation programs that target those most at risk (Ding and Wei 2022; Donald 2014). In order to proactively prepare for drought and increase options for drought adaptation, early warning information is required. This can be delivered through a user-centered integrated Early Warning System (EWS) for drought, where all components of such system are intrinsically interconnected (Zia and Wagner 2015; Bhardwaj et al. 2021a). If an EWS provides a staged warning of drought conditions, decision-makers can take anticipatory actions to mitigate potential damages and even reduce vulnerabilities (Andersson et al. 2020).

EWSs can be described as a set of capacities needed to generate and disseminate timely and actionable warning information to assist with proactive decision-making. This enables individuals, communities and organizations at risk to a hazard be prepared, reduce their vulnerability, and act appropriately reducing the degree of harm or loss (Kelman and Glantz, 2014). The majority of current EWS applications are focused on rapid-onset hazards such as floods, tropical cyclones, tsunamis, and earthquakes. On the other hand, EWS infrastructure for slow-onset disasters like drought is comparatively less advanced (Banholzer et al. 2014).

There are four fundamental areas within an EWS: risk knowledge, monitoring and analysis, communication and dissemination, and response capability. Risk knowledge refers to analyzing data related to risk that is systematically collected. Monitoring and analysis require reliable technology and involves detection of hazards and hazard impact monitoring. Response capability refers to educating vulnerable actors through preparedness programs and providing support and aid (Sai et al. 2018). The third and historically weakest element of an EWS, communication and dissemination, is the focus of this research.

Effective communication involves using clear, succinct and understandable messages disseminated in a timely manner that can engage, empower, and prepare users for future impacts (Harrison et al. 2020). While all components of an EWS can be further improved to increase efficacy and usability, it is widely acknowledged that an EWS is only as strong as its weakest component. Earlier studies show that communication and dissemination is a weak and overlooked element of EWSs, making it a main area for improvement of overall EWS impact and actionability (Emma 2003; Sufri et al. 2020b).

The communication infrastructure for EWSs has to fulfil three primary tasks: (i) sensor communication, i.e., measurement data sent to the warning center(s) from the ‘field’; (ii) center-to-center communication, i.e., direct communication between experts, warning centers and/or development and testing sites, and (iii) communication to users, i.e., delivery warning and alarms to the general public (Chaves and De Cola 2017). The sensor communication element is commonly used for EWSs where the monitoring and forecasting information is crowdsourced, e.g., local community members sharing soil moisture observations. In the case of an EWS for drought where remote sensing systems are used and received data is stored in a central repository, this communication element would not be included in the communication and dissemination component.

There is an extensive range of communication technologies that can be used to disseminate warning messages to at-risk populations. Each communication technology has different efficacy characteristics and thus has different impact and success (Chaves and Niebla 2014; Hou et al. 2024). The efficiency of the transmitted warning messages is affected by both the organization providing the information, communication technology used and the corresponding receiver device. Therefore, it is important to build trust and understand the needs of those being communicated to when developing the communication practices used to deliver drought risk information.

It is unknown, particularly within an Australian context, whether the drought information that is available includes early warning information, or whether it is being communicated effectively. There is a diverse amount of literature on how to make communication effective for EWSs for drought but there is limited assessment on whether such strategies are being successfully implemented and utilized for drought communication in the MDB. Given that current MDB drought communication is varied and sporadic, it is evident that the region would benefit from the creation of a general drought communication strategy that can be localized as required.

2 Data and method

The study used a qualitative, social science approach involving semi-structured interviews in three states of the MDB - New South Wales (NSW), Queensland (QLD) and Victoria (VIC) (Fig. 1). Within each state, multiple Local Government Areas (LGAs) were assessed including Goondiwindi Council and Western Downs Regional Council in QLD, Gannawarra Shire, Mitchell Shire Council, Murrindindi Shire Council, Loddon Shire Council, Buloke Shire Council, Yarriambiack Shire Council, Rural City of Mildura and Swan Hill Rural Council in VIC and Wagga Wagga City Council and Gwydir Shire Council in NSW.

Fig. 1
figure 1

Map of the study area - the Murray-Darling Basin, Australia

Full size image

For this research, farming individuals were selected based on the following criteria: (i) owned, consult on, or managed a farm that was located in the MDB in VIC, NSW, or QLD; and (ii) had experienced drought before. The participants, male and female farmers, varied in age between 28 and 74; their farms varied in region and type of farming operation. This allowed for the understanding of how different demographics affected an individual’s communication needs. This study was not concerned with the effect of specific locations on interviewee response. Rather, since the MDB is a heterogeneous farming region, the arbitrary and diverse LGA selections allow for high-level comparative analyses between regions and demographics. A demographic overview of each LGA area was completed using the Australian Bureau of Statistics (ABS) to further understand the complex and diverse populations in different locations (Table 1). This was used to analyze key trends between participant demographics.

For this research, farming individuals were selected based on the following criteria: (i) owned, consult on, or managed a farm that was located in the MDB in VIC, NSW, or QLD; and (ii) had experienced drought before. The participants, male and female farmers, varied in age between 28 and 74; their farms varied in region and type of farming operation. This allowed for the understanding of how different demographics affected an individual’s communication needs. This study was not concerned with the effect of specific locations on interviewee response. Rather, since the MDB is a heterogeneous farming region, the arbitrary and diverse LGA selections allow for high-level comparative analyses between regions and demographics. A demographic overview of each LGA area was completed using the Australian Bureau of Statistics (ABS) to further understand the complex and diverse populations in different locations (Table 1). This was used to analyze key trends between participant demographics.

Table 1 Demographic overview of LGAs.
Full size table

There is a range of populations, with Gwydir Shire Council having the lowest total population, at 5,353 people and the Rural City of Mildura having the highest population, at 55,777 people. The gender split is mostly 50% male and 50% female for all LGAs analyzed. Loddon Shire Council has the most agricultural, forestry and fishing sectors, with 65% of businesses from that sector. Mitchell Shire Council has the lowest amount of that sector, with 12%.

Statistical data collected from the farming individuals interviewed are presented in Table 2. The age varied, from 20% of participants aged between 23 and 35, to 40% aged 65 and over. Gender-wise, those who identified as males made up about three quarters of participants, compared to one quarter of participants who identified as females. Majority of participating farmers (73%) were from the LGAs in VIC.

Table 2 Demographic overview of participating farming individuals
Full size table

Semi-structured interviews allowed for exploring participants’ experiences with past drought events. A semi-structured interview guide was developed prior to the interviews and used to allow for a greater depth of data to be gathered under the phenomenological research method (Ruslin et al. 2022). The semi-structured nature also allowed for contextual adjustment while minimizing the risk of receiving incomplete answers which may sometimes be the case for data harvested through a structured interview or survey (Adeoye-Olatunde et al. 2021; Longhurst 2003). The interview guide was divided into three sections: (i) Identifying demographics, (ii) current drought practices, and (iii) early warning systems. The broad topics of interview questions were created based on recommendations of earlier studies (Bhardwaj et al. 2021a).

The interviews were conducted over the phone in April – June 2021 (covid-related travel restrictions prevented conducting face-to-face interviews). The audio of the interviews was recorded using ‘Otter Voice Meeting Notes (Otter.ai) Version 2.3.73’ to partially transcribe the recordings and then further transcribed using the online Otter Voice Meeting Notes (Otteer.ai) capabilities. A PDF of the transcripts for each de-identified interview was coded and analyzed using NVivo – Version 20 (release 1.4.1) which is an industry standard for qualitative data processing (Austin and Sutton 2014).

The data analysis method chosen was thematic analysis, as it allows for the identification, analysis, organization and description of themes found within a data set (Nowell et al. 2017). An inductive approach was taken to generate conclusions from the data collected by moving from specific farmer’s experiences to broader generalizations on communication needs. Themes were drawn out from the data rather than from previously identified themes established in the literature (Naeem et al. 2023; Tracy 2010). Data was consolidated to (i) identify potential themes, (ii) code common words, key statements, phrases, or patterns, and (iii) review potential themes and compare them against the overall dataset to ensure the themes were representative of the data.

3 Results

The interviews were analyzed to understand farmers’ experiences and perspectives with (i) the current drought communication strategies, and (ii) preferences for future information dissemination. Three themes were found to explain the current drought practices used by participants: current communication channels, relevance, and trust in current communication. Furthermore, three themes were found to explain the future considerations for early warning information for drought: future communication channels, community, and trust for future communication. These six themes were used to characterize farming individuals’ communication needs and barriers to the current drought information and their preferences for future early warning information for drought.

3.1 Perspectives on current drought communication strategies

3.1.1 Communication channels

Existing drought ‘communication channels’ were mentioned by all farming individuals and was identified as a main theme. Farmers had a variety of sources of information in regard to drought risks, monitoring and forecasting. The current way most participants find their information is through online mediums, with 73% using webpages, e.g., the Bureau of Meteorology website, Weather Zone etc., 27% using weather-based YouTube videos, 40% using the Department of Agriculture’s Fast Break online newsletter and 33% using mobile phone apps. Seminars from professionals were also sources of information for 13% of participants. 27% of participants use newspapers and television programs and 13% listen to information from the radio to obtain drought information. Every farming individual used more than one communication channel to obtain their drought information.

40% of participants spoke on the digestibility of information that is provided before, during and after drought, e.g., commenting that they have to rely on the written text or the description provided by a forecaster. Easy to understand information is a motivator for farmers to use the drought information provided. As well as this, 33% of participants mentioned that the accessibility of information also was a motivator to using drought information.

3.1.2 Relevance of information

For current drought communication, 87% of interviewed participants spoke about the ‘relevance’ of information. As most stated that information was general and not location-specific, it was identified that this was a barrier to the information being used effectively. Some participants spoke on the importance of communicating with other farming individuals in their area to obtain more relevant data. Accuracy and reliability were other factors mentioned when talking about the relevance of information; 67% of participants mentioned accuracy of information throughout their interviews and noted it to be a motivating factor. This indicates that increased accuracy is likely to lead to increased trust and actionability.

In cases of long-range climate prediction where model skill is typically reduced throughout the months and seasons, and will never be completely accurate, warning information should include such context to users. For example, if a drought “watch” is issued using an early season forecast with reduced skill, the “watch” should include a statement from a climate expert that clarifies users to be alert but aware that model skill may evolve through the course of the season and result to a differing drought status. Alternatively, in warning transitions, with high skill and high urgency (i.e., a transition into the highest level of drought warning – “drought emergency” – with high accuracy input forecast), warnings should similarly communicate the elevated risk and high confidence in the warning as this extra information could further emphasize the need for urgent preparation.

3.1.3 Trust in current communication

“Trust” and reliance on provided information was mentioned by 93% of participating farming individuals. It was found that interviewees are not likely to base their current decision-making on drought information alone. With the majority of participants stating that they only use forecasts and drought information as one element in a diverse dashboard of information informing their decision-making.

The majority of participants spoke on trust in terms of the information itself. However, few mentioned trust in the information provider. Interviewed farmers highlighted that quality and quantity of information are significant factors that influence trust. This emphasizes the need for long term communication that priorities transparency.

3.2 Perspectives on future drought information dissemination

3.2.1 Communication for EWSs

All participants expressed their needs and preferences for future ‘communication channels’ to deliver early warning information. Participants were interested in the ability to have early warning information, as 87% stated that current information that they use does not involve any early warnings for drought.

There was no consensus about particular communication channel that best suit an EWS, as participants from the older age groups (55–65 and 65+) were interested in receiving information from radio, television, and newspapers while farmers from younger age groups (25–35, 35–45 and 45–55) were interested in mobile applications, text messages and emails. This disparity across demographics highlights the need for warning information to be delivered in multiple ways and through multiple mediums. Interviewees also emphasized the importance of EWS to disseminate information that is ‘clear, understandable and useable for all.’

Participants went on to emphasize the time-poor nature of farming and highlighted that a system that could be used and accessed easily and quickly would motivate many to use early warning information.

3.2.2 Community

‘Community’ was a reoccurring theme in 53% of interviews as a way of improving future drought communication through EWSs. It was suggested that local knowledge could play a role in these systems. Many participants already utilize the community around them and share information amongst neighboring farmers, and therefore would find the early waring information better that it comes from people who have experience with farming, or at least have farmers input into the information that will be provided. This notion is also prevalent in the literature with some studies evaluating the importance of ‘local champions’ in disseminating trusted local warnings (Andersson et al. 2020; Sufri et al. 2020a). Commenting on role of local government officials, participants indicated that such officials currently play a minimal role within creating, using, or maintaining an EWS for drought. However, some participants noted that LGA newsletters and announcements could be a viable localized source for drought information.

3.2.3 Trust in future communication

All interviewees mentioned trust for future communication of early warning information. About half of the participants would be willing to rely on early warning information more than the current information. Accuracy was revealed again within this theme as a key driver that influences trust in future drought communication. Building a trusted relationship with information providers was also important to farmers, who would prefer information to come from sources they are already familiar with. Many commented that an EWS would be used as another tool amongst a multitude of other tools and knowledge they use to make drought-related decisions. It was emphasized that the system would have to be trialed and tested to demonstrate its reliability before farmers would be willing to trust it. This identification of a system’s hit and miss rate is common in climatological operationalization and the obtained results highlight the importance of such verification studies.

4 Discussion

The aim of this research was to explore the drought communication needs of farming individuals through their perspectives and experiences. This section brings together the core emergent themes of the interviews to discuss the most significant needs and consideration for organizations attempting to provide effective communication of drought information.

4.1 Communication channels for current drought information

In Australia and globally, agricultural information studies traditionally have evaluated farmer’s information source preference. The novelty of this research is its ability to highlight the channels Australian farmers currently use, their experience using them and why they use them. The results showed that most farmers access their drought information from online sources. The way in which agricultural workers access information has greatly changed over the past decades, due to the advancement of technology and digitisation of the industries involved. As indicated by Starasts (2015), this has been complemented by a substantial increase of online agricultural content.

The interviewed farmers were aware of the vast amount of knowledge available; however, many found the information itself difficult to access and digest. Some farmers stated that the online webpages that they commonly used were difficult to navigate. Drought information sources often used technical and scientific language that without appropriate training many noted to be difficult to comprehend.

In the past, climatologists from the National Climate Centre at the Bureau of Meteorology visited rural communities participating in agricultural field days, presenting seasonal climate outlooks, clarifying farmers’ questions about drought information etc. However, traditional person-focused targeted approaches to educating and training farmers are in decline in Australia (Hunt et al., 2012; Sewell et al. 2017). These days, to obtain drought information, famers should more often rely on various online sources. Another difficulty found with the large number of sources available is that different sources can provide conflicting information.

Information access is limited by rural telecommunication and internet infrastructure. This means that many farmers continue to use newspapers, television, and radio services to obtain information. Exclusively digital information runs the risk of bypassing remote areas that are not covered by mobile phone networks, creating an inequality of access to data.

The farming individuals emphasized when seeking information, they use a multi-channel approach where they obtain drought information from multiple sources. While farmers consider such information, they use own intuition to make decisions for their property. Farmers are facing a challenging task to find and then critically evaluate a large volume of drought information from many, often unrelated digital and non-digital sources to make decisions appropriate for their property (Melzer et al. 2023; Sørensen et al. 2010) in this way building resilience and proactively managing drought events.

4.2 Relevance of current information

Currently, a significant challenge for decision-makers is communicating reliable and appropriate drought risk and proactive strategies on a local scale (Kelman and Glantz, 2014). Farmers interviewed expressed concern about existing gap between the currently available scientific drought information which was perceived as too broad and the information that they considered as useful in their day to day activities. This is in line with finding of Kiem and Austin (2013) about the complexity of farmers’ needs in adjusting general scientific information to a specific location. Farmers identified that, while there are numerous sources of drought information, there was limited localized information surrounding drought and described the current drought information as general. Drought can look very different from one farm to another, even within the same LGA, due to its dynamic nature. Typically, those who did find the available information to be reasonably region-specific were in larger LGAs. This suggests that there may be less of a focus on generating region-specific information for smaller LGAs, such as Gwydir Shire Council in NSW and Buloke and Yarriambiack Shire Council in VIC.

Farmers also recognized that they required accurate information for them to use the drought information effectively but were not receiving accuracy for their specific locations. Timely and accurate climatic and meteorological data is critical for farmers’ decision-making, with many placing their livelihoods on such information. There was understanding amongst some farmers that accuracy of climate models decreases further into future predictions.

Results showed many farmers were interested in increasing the transparency of why certain forecasts were inaccurate and wanted reliable communication after a forecast did not come to fruition. Solving this requires a structural shift. Fakhruddin et al. (2020) showed that there needs to be public-interest media with viable business models, better interest governance, education systems that support media literacy and critical thinking skills, and political leaders and other influencers that condone sharing inaccurate and misleading information instead of perpetuating it. On an individual level, experts and those who work in the field should be conscious when sharing complex content on social media and within their communities.

In order for drought information to be effective, communication should focus on the people exposed to risk and their perception and knowledge of what that risk means. This is also true when providing different sectors with specific information. For example, case studies in Europe indicated that sector-specific needs in terms of drought vulnerability were important in identifying appropriate risk and management strategies (Blauhut et al. 2016; Kelman and Glantz, 2014). A water supply company may want detailed hydrological information to forecast public water supplies. However, farmers may require advice about irrigating crops. Farming individuals not only need different information compared to the general public but need specific information that complements the type of farm they run. The livestock and crop farmers interviewed highlighted their differing needs in terms of timing of when information is disseminated and the content that is provided. This process is elaborate due to the complexity of social and institutional mechanisms and the multiple potential sources of climate and drought information, not all of which is consistent or easy to understand. An earlier study by Starasts (2015) which supports the findings from this research showed that when farmers are seeking information to make decisions, a strong sense of their specific and individual situation guided and defined the methods of findings appropriate information. Creating relevant drought information can be achieved by including education, awareness, and continual data collection so that it becomes familiar to and accepted by the community and each sector. Relevance is essential to providing accurate and reliable monitoring and predictions as it allows for information to be tailored to the needs, attitudes, and vulnerabilities of at-risk individuals and communities.

4.3 Trust in current communication

Farmers stated that they were willing to consider the current drought information, but not willing to solely rely on it. Many farmers would use the current information as just another tool and will easily dismiss an organization’s information if it was believed to be unreliable and inaccurate. It was suggested in earlier studies that the extent to which an organization, e.g., disaster management office, manages drought impacts (i.e., impacts on social infrastructure) will affect trust in the organization (Badu et al. 2023; Kramer and Carnevale 2003; Reynolds and Quinn 2008). The way in which drought information providers engage with the communities (i.e., the quality and quantity of communication) and treat community members (i.e., procedural fairness) shapes community trust in information providers and thus acceptance of the drought information it provides (Moffat and Zhang 2014).

Trust is also shaped by past personal experience (Weiss et al. 2020). If farmers made a costly decision based on low skill forecasts then they will likely be negatively charged towards future forecasts, even if the forecasts are of high skill. So prioritizing trust would involve both optimizing the system as well communicating with empathy and transparency about changes in forecasting skill.

The basis for enhanced decision-making is availability of timely, high-quality data. However, the current situation in farming is that most data and information sources are difficult and time-consuming to use. This indicates that the full potential of such drought data and information are not being fully exploited (Fountas et al. 2015). All farmers interviewed stated that they looked at weather and climate information; however, the specific timing varied, e.g., typically, farmers checked weather forecast daily, an extended 7-day weather forecast - two to three times a week, and a seasonal climate outlook - once or twice a month. The difficulties in looking for current information, and the day-to-day commitments farmers must attend to, appear to hinder their ability to access and understand complex information (Hunt et al., 2012). Therefore, the timeliness in which information can be digested and is disseminated heavily influences the usability and trust of that information (Krishna and Sivapuram, 2022).

A majority of farmers interviewed sought out familiar sources for information on drought, e.g., the Bureau of Meteorology which farmers had previous relationships with and already trusted due to the organisation’s credibility and reliability of information provided. Organizations that provided face-to-face seminars, where they revisited farmers and repeated information, built trust with farmers, especially with new types of information or recommended actions, such as forecasts. A close and ongoing relationship between data providing agencies and users was found to be a contributing factor in trusting drought information. This was also found by Andersson et al. (2020) and Starasts (2015) who uncovered that more rural communities would learn in isolation of their local communities. Due to the dynamic nature of drought, with farmers mentioning that ‘no two years in farming look the same’, it is important that trust is there in both the information provided and the agency that is providing such information.

4.4 Communication channels for EWSs

The main aim of an EWS is to prevent hazards from becoming disasters. An EWS, however, can only protect those at risk when a warning is received in a timely fashion, the meaning of the warning is understood, and the appropriate actions are taken (Fakhruddin et al. 2020). Therefore, the communication channels play a critical role in an EWS for drought.

From the interviews, there was no consensus on a specific preferable communication channel through which early warnings should be disseminated. It was made clear that information that was relative, trusted, and easy to understand was more important. A multi-channel approach would be appropriate to use in order to reach those at risk, and the channels should be familiar and user-centered in order to cater to needs of all individuals. The integration of spatial and temporal historical data, real-time farm data, knowledge sources, statutory compliance, health and safety guidelines, environmental guidelines, economic models, etc., into a coherent management information system is expected to remedy this situation as well (Fountas et al. 2015).

Much can be learned from EWS for rapid-onset hazards such as floods, storms, etc., which are already used by some of the interviewed farmers. For example, dissemination and communication element of a flood EWS consists of three aspects, which are: (i) organizational and decision-making processes, (ii) effective communication systems and equipment installed, and (iii) warning messages recognized and understood (Werritty et al. 2007). The timing of these messages is crucial, with most communication strategies for floods outlining that early warning information should be disseminated for a period of more than seven days before the event occurs.

The novelty of this research demonstrates the difference that drought makes to the communication and dissemination of early warning information, as drought has a ‘creeping phenomenon’ and dynamic nature. This is a distinguishing difference as this slow-onset hazard changes the communication needs and the perception of risk. It is necessary for the information to be provided more further in advanced.

Farmers also supported the need for early warning information to be available a few months in advanced to enable them to make proactive decisions on their property. As demonstrated by Bhardwaj et al. (2021b) in a case study for PNG, a contemporary user-centered integrated EWS for drought has capability to provide accurate early warnings 3–5 months in advance. Implementing such modern EWS for drought in Australia is recommended.

In drought risk reduction, there is often an emphasis on planning and foresight to ensure reduced impact, yet risk communication is frequently at the tail end of that process, rather than integrated throughout (Mysiak et al. 2016). As humans are complex, cognitive processes can lead them to disregard what may be sound information and advice. This indicates that communication must become a process, not a product, in order to improve risk communication.

4.5 Community

EWSs for drought are a key factor in improving outcomes for affected communities (Pulwarty and Sivakumar 2014). Community engagement is maximized by making systems user-centered, which address the human and socio-economic factor in EWSs for drought (Basher 2006; Pham et al. 2024). Farmers acknowledged that relevance and accuracy was missing from the current drought communication, and that knowledge from locals could improve this.

Despite this evidence that community input is desired and is a key factor in success of an EWS, it is not always commonly sought in many natural hazard assessments, policy development and implementation processes (Sufri et al. 2020a). Diverse community knowledge can be used as a bridge in EWS for drought to allow such systems to be region-specific by providing local knowledge of the key vulnerabilities to a hazard. Meadows et al. (2014) found that by utilizing education lead farmers, who can engage with other agricultural stakeholders, effective drought communication that is relevant to farmers can be achieved. It was important to many farmers that their voices were integral parts to the planning, creating, and disseminating of drought information as the goals of the farming community can be shifted relatively rapidly that larger agencies may be unable to keep pace with. This can be accomplished through active participation of the population in monitoring, mapping risks and through transmission of information between communities and information providers.

Community participation allows for the combination of technological advances and technical knowledge, and local monitoring in order to create information that is relevant and specific to particular areas (Zambrano et al. 2017). An earlier study exploring the adoption of agricultural communication found that access to information was key to uptake, and that peer-to-peer knowledge sharing was more effective to generate support than information from authoritative sources (Varble et al. 2016). Before the implementation of an EWS for drought, it would be beneficial to create a feedback system which allows for local farmers to input into the system.

Utilizing the knowledge from more experienced farmers was highlighted to help create user-centered EWSs. In spite of this, rural community resilience can be impacted upon by regional institutions including local government (Eachus 2014). Therefore, in decision-making, local authorities should consult an appropriate range of stakeholders to avoid flawed decisions damaging wider community resilience (Henderson et al. 2020; Werritty et al. 2007). However, many farmers believe local government plays a minimal role in creating, using, and maintaining an EWS for drought. Nonetheless, some participants noted that, through LGA newsletters and announcements, drought data and warnings could be disseminated via the local government for the information to be region-specific. Research into the use of local radio, newspapers and community media to disseminate information has been discovered to play a crucial and often underappreciated role such community members fulfil at times of hazards such as drought. These forms of information sources provide the type of news community members consider highly reliable and can also support natural disaster planning and response (Ewart 2020). This information would need to be sourced and referenced from sources that the farmers already used to begin to build rapport with the farmers. Where a hazard like drought is repeatedly experienced, it becomes increasingly important that local agencies provide leadership in drought mitigation and management to maintain community resilience. EWSs for drought can be stakeholder informed and locally disseminated but require a large amount of community engagement, participation, and trust for the system to be effective.

4.6 Trust for future communication

Trust in information can be a strong indicator of community acceptance of an EWS for drought. Many farmers were interested in obtaining early warning information, as their current information does not include any early warnings and were willing to rely on it more than the current drought information. There were, however, points made surrounding the factors that would influence their trust in the system. Accuracy of information was highlighted as a major factor that would influence whether farmers would use early warning information for drought. Before disseminating information to agricultural communities, trials will need to be completed for the EWS for drought to obtain assurance that it is mostly accurate in order for farmers to intake the information effectively. While accuracy has been recognized as an element that can influence trust, it is important to make clear the true meaning of the information that is being provided. In terms of the monitoring and forecasting, users need to understand that they are being provided with probabilistic information, e.g., chances of rainfall to be above or below normal, as well as with background information about major climate drivers and their expected influence. This can be done through public outreach and education, which may be critical to building trust in the EWS for drought and its provider to support and increase the uptake of the information. Often, misinterpretation of early warning information can create mistrust and apprehension towards the organizations releasing early warnings (Buchanan-Smith 2000; Kuller et al. 2021; Wilhite 2014). Therefore, including members of the public who are not climate experts in the conception of how the EWS is communicated, and the content of the warning, can act as an effective mitigation strategy for the potential misinterpretation of information for end users (Buchanan-Smith 2000; Siegrist 2019).

Another factor that would motivate farmers to trust information from an EWS for drought was the relationship between the user and provider. Research shows that the reputation of organization that implements the EWS for drought and disseminates the warning information is crucial to the rate of trust and use by stakeholders (Kochanek et al. 2020). From the interviews, many identified that it did not specifically matter who the provider of the EWS for drought is, as long as an agency is reputable, known and already trusted by users. Building a trusted relationship is vital for the intake of information. Users of an EWS are more likely to engage with messengers they trust. Any perceived dishonesty can abolish that trust immediately, and future messages may not be respected and may become less valued (Bokal and Muller 2018). Maximizing the value of connectivity between providers and users to support farmer’s transition to more proactive drought measures, and to improve trust in the future use of early warning information in agricultural development, is a key challenge for EWS developers.

4.7 Demographic and social environmental factors

In the MDB, drought is just one of the many challenges facing local farmers, with the broader shifts in the farming sector associated with the changing global markets and rural demographics (Kiem and Austin 2013). The interviews found that particular demographics correlated with certain experiences. While correlation between demographics and communication needs cannot be statistically established, there were trends observed within the answers given to interview questions. Farming individuals from older age ranges tended to use less web-based applications and online sources, and preferred to receive information from newspapers, radio and the community around them. Those who were aged 65 + made up 40% of the participants and it was observed that their extensive experience in farming allowed them to be confident in relying much more on their own intuition rather than the provided drought information. Younger farmers seemed to use online sources and agronomists to obtain their drought information and were relatively more open to seek new information from different sources. Ensuring that a localized user-centered EWS assists all demographics of a community can be challenging, considering different ages, genders, abilities, religious, ethnic, and cultural groups. Despite this array of factors, it is evident that the communication and dissemination of drought early warnings needs to appeal and cater to all needs and abilities (McEwen et al. 2020).

It is important to make note of the social environmental factors and unique lifestyle of rural perspectives. Australia has one of the most highly concentrated, unevenly distributed and, for its land area, the smallest populations. The resulting large disparity in density and extensive areas of sparsity make servicing rural and remote areas of the continent difficult (Karacsonyi and Taylor 2023; Vinson 2007). Partially because of this, Australian rural communities are vulnerable to agricultural risks, as well as the socio-economic impacts, of drought because because these areas depend on water-intensive industries and primary production, in addition to the effects related with accessing health care and educational advantages (Sullivan et al. 2018; Austin et al. 2018).

The difficulties rural communities face can impact on the way in which they prioritize and understand information. Interviewed farmers highlighted the difference between the information required by those in urban areas compared to their needs in rural contexts, as in their decision-making farming communities depend comprehensively on the weather and climate information. Farming individuals rely heavily on such information due to the pressure on agricultural industries to constantly produce commodities in changing, and often hostile, climate. This makes it even more vital to comprehend how rural and regional communities and individual perceive and understand risk. If the user of an EWS for drought is to be put first, the way in which users understand and interpret information needs to be known for information providers to communicate effectively.

The communication of drought early warnings needs to be catered to the way in which farming individuals take in information. Tailoring information to adjust for various educational levels is vital to ensure that early warnings are understood and actionable. Fielke and Bardsley 2014; stated that ‘education increases an individual’s abilities to employ and embrace new technology and holistically envisage their agricultural businesses’. Formal farmer education programs supported by Government, as well as the encouragement for further technical and community-based learning and monetary investment for farmers would help incorporate positive social, economic, and environmental outcomes within the agricultural sector in the MDB (Fielke and Bardsley 2014). It is particularly important that such investments are not short-term media displays of reaction and instead long-term commitments in both drought and non-drought times.

4.8 Further research

The research conducted contributes to developing the foundation upon which a communication strategy of a user-centered integrated EWS for drought is built. Further work should seek to broaden the scope of communication and dissemination of drought information. In particular, this study is focused on the dissemination itself, but a greater focus could go into the information and content that is being delivered, what it is and how that is presented. This could also provide a deeper emphasis on how location, rural and urban contexts affect understanding of drought information. It is recommended to further evaluate communication strategies engaging other stakeholders, e.g., agronomists, insurance services and information media organizations. This would provide further insight into how various stakeholders perceive the current drought communication and highlight their influence on future early warning information for drought.

Ultimately, shifting the focus from reactive to proactive measures through drought mitigation, vulnerability reduction and preparedness is vital in the changing climate. This can be done through developing and implementing a user-centered integrated EWS for drought in Australia, which is of interest to farming individuals and would be utilized if shown to be accurate, accessible and is communicated effectively. Future work could seek to promote the evolution of drought knowledge bases and to establish a mechanism for sharing knowledge and providing services to stakeholders at different levels, creating a two-way dialogue with farming communities to input local knowledge.

5 Conclusions

This research identifies the communication needs and barriers of drought information for farming communities in the Murray Darling Basin, Australia. It was explored by identifying the effectiveness of the current drought communication practices used by farming individuals and the barriers to access and use the information. To investigate how early warning communication for drought could best suit farmers in the agricultural industry that work within the MDB, interviews with farmers in Victoria, New South Wales and Queensland were conducted. Participants were asked about their current experience with using drought information as well as projected needs for future communication in the form of an EWS for drought, which is currently being developed.

Six emerging themes were found within the interviews with farming individuals, including current communication channels, relevance, trust in current communication, future communication channels, community, and trust in future communication. The results showed that the current communication channels used varied amongst farmers and the ability for farmers to access, understand and use drought information is vital for designing future EWSs. A significant challenge in the current information is communicating relevant, accurate and appropriate drought risks. For future EWS, utilizing the community and local knowledge could aid in creating tailored information that is specific to locations and individuals.

Trust was a major factor in both the current drought communication and in developing the communication of EWSs. Information providers should focus on building productive relationships with users and creating a system that is transparent in order for it to be effective for farmers. When developing drought communication, understanding the way in which users will interpret information, as well as the context in which they operate under, needs to be known to create a user-centered experience.

Drought will continue to be a pervasive part of the Australian environment, and its frequency and severity will likely increase as the climate changes. It is therefore of the highest importance that the way in which drought information is communicated benefits those at greatest risk to drought by prioritizing a user’s needs and experiences in all stages of drought EWS development. Despite these findings being centered on primary producers in the MDB, the in-depth user-engagement scoped in this study provides insight into efforts to develop a user-centered integrated EWS for drought at the national level.