1 Introduction

Floods are devastating natural hazards with widespread impact, causing significant economic and social damages across the globe (Benson & Clay, 2002; DMSG, 2001). Many regions in Asia, including parts of India and Bangladesh, face recurring severe floods during the monsoon season, leading to extensive damage and loss of life (BBS, 2017). The area in the northeast part of Bangladesh, commonly referred to as the “Wetland Basin,” is particularly vulnerable to flash floods, making it one of the main natural disaster-prone areas in the country (Irfanullah et al., 2011; Khan et al., 2012; Miah, 2013). Beginning with the monsoon season, the Wetland Basin is characterized by its low-lying river basin geography, which remains submerged for approximately half the year (NERP, 1995). This prolonged inundation, lasting from June to November, poses significant challenges for agricultural activities, particularly Boro rice cultivation (CDMP-II, 2014). Boro rice is the primary rice variety cultivated in Bangladesh, contributing to approximately 55% of the rice produced (Shahid, 2011; BBS, 2017). The flash flood-affected region accounts for about 20–25% of the annual Boro rice production, highlighting its critical importance in local, regional, and national food security (BBS, 2017; Huq et al., 2012; Rabby et al., 2011).

Assam and Meghalaya, India, are experiencing rapid rainfall-runoff, leading to flash floods in the Wetland Basin. It is weeks before the Boro rice harvest when the situation threatens the standing crop (Ahmed, 2012). The occurrence rate of flash floods before the monsoon season (March to April) has been reported to impact the region several times in the past few years (Biswas, 2017; Islam, 2015; Roy et al., 2017). These flash floods result in substantial damage to the Boro rice crop, threatening the staple food supply and impeding the overall progress and development of the country, as the wetland region constitutes a significant portion of Bangladesh (Haor Master Plan, 2012). Besides the local challenges, Bangladesh, which ranks 8th in the world and has a large population, faces the broader challenge of ensuring agricultural production and food security due to adverse climate change (Transforming Our World: 2030 Agenda for Sustainable Development). Climate change has diverse manifestations across the globe, particularly affecting natural resources and the agricultural sector (Lobell et al., 2008). The impact of uncertain events, exacerbated by global warming, has resulted in substantial losses in Boro rice production, threatening food security and posing a significant challenge to sustainable development goals (SDGs).

A comprehensive understanding of economics, environmental science, and technological advancements is necessary to fully understand the nexus between climate change, coping mechanisms, and Boro rice production in Bangladesh’s wetland areas. Recent research has emphasized the effects of climate change on wetland ecosystems, encompassing alterations in precipitation patterns, increasing temperatures, and extreme weather occurrences (Akter et al., 2023; Mahtab et al., 2018). These changes significantly affect wetland agriculture, which relies on wetlands’ natural water regimes and ecosystem services.

To address this issue and explore possible solutions, researchers and policymakers have started to apply a combination of economic analysis and satellite remote sensing technology (Islam et al., 2021a, 2021b, 2022a, 2022b; Kamruzzaman & Shaw, 2018). This integrated approach allows for a better understanding of the economic dynamics and environmental impact associated with wetland agriculture, as well as the development of coping mechanisms and adaptation strategies. The economic approach plays a crucial role in analyzing the costs and benefits of wetland agriculture and its vulnerability to climate change. Cost–benefit analysis can help assess the economic viability of wetland farming practices and guide decision-making processes. It can also inform the design of economic instruments, such as subsidies or insurance schemes, to incentivize sustainable wetland management and promote climate-resilient agricultural practices. The use of satellite technology, such as geographic information systems (GIS) and remote sensing (RS), can provide valuable data for monitoring and assessing changes in wetland ecosystems (Ahmed et al., 2017; Nahar et al., 2017). Satellites can capture information on land cover, vegetation dynamics, hydrological patterns, and climate variables, which are essential for understanding climate change's effects on agriculture in wetlands areas globally.

Given the critical importance of Boro rice and the increasing vulnerability of the wetland areas to flash floods and climate change, it is essential to explore practical strategies and coping mechanisms to mitigate the adverse impacts on agricultural productivity (Jakariya & Islam, 2017; Kamal et al., 2018). Hence, addressing these challenges is vital for achieving SDGs and ensuring the long-term resilience of wetland agriculture in Bangladesh. This study investigates the relationship between Boro rice production and climate change through advanced satellite remote sensing systems along with economic approaches. Furthermore, alternative probable coping strategies, including land suitability assessment, vulnerability mapping, yield forecasting models, and damage-based crop insurance techniques, were discussed to ensure sustainable wetland production and secure food supplies in Bangladesh.

The entire manuscript is structured as follows: Sect. 2 discusses the literature on wetland agriculture in Bangladesh. Section 3 provides an overview of the country’s current state of wetland agriculture. Section 4 examines the impacts of climate change on wetland areas in Bangladesh. The effects of flash floods on wetland areas are detailed in Sect. 5. Section 6 examines the vegetation phenology of wetland Boro rice production using satellite data. Section 7 highlights probable strategies to mitigate the effects of climate change. Section 8 identifies challenges and proposes potential solutions. Finally, Sect. 9 concludes the study and offers policy recommendations.

2 Literature review

The study considers published research articles from two renowned databases, including Scopus and Web of Science (WoS), for a literature search. Multidisciplinary academic databases cover scholarly literature. Initially, various keyword combinations about the topic under investigation were utilized without specifying a particular discipline. Subsequently, the search was narrowed down to the field of social sciences, and specific journals were chosen. The selection of journals aimed to gather research findings from diverse sources while excluding those focused on medicine and aggression. The following keywords, “haor OR wetland AND Bangladesh AND climate change,” were used to achieve the maximum outcome. These terms were incorporated within the articles’ title, abstract, or keywords. The search was confined to English-language studies. Network analysis, a method presenting interactions among various computational attributes graphically, is facilitated by diverse resources. The visual representations (Fig. 1) utilized in this study were generated using VOSviewer software, illustrating network analysis diagrams derived from a combination of diverse computable parameters extracted from the WoS (Fig. 1a) and Scopus (Fig. 1b) databases. VOSviewer software is freely accessible for download. This tool assesses bibliometric networks based on these parameters, utilizing input files with.txt extension from WoS and.csv extension from Scopus (Mulay et al., 2020; Kadam et al., 2016).

Fig. 1
figure 1

Source: Authors’ own development

a and b Visualization of common co-occurring keywords based on a WoS and b Scopus articles of wetland areas of Bangladesh and climate change.

Full size image

Three types of visualization techniques are accessible: networks, overlays, and densities. Figure 1a and b exhibit a network visualization diagram generated from a blend of keywords and source titles sourced from both WoS and Scopus. Circles on the map represent the keywords extracted from the titles of the sourced documents. Larger circles indicate more frequent use of the keyword, with smaller circles indicating less frequent use. Connections between circles signify the proximity between the respective keywords, with the strength of the association reflected in the size of the connecting lines. Keywords sharing similar colors denote groups of closely interconnected terms. Both diagrams have three clusters, and a distinct hue represents each one.

Numerous researchers have undertaken investigations in recent years aimed at finding sustainable solutions. For example, studies have focused on livelihood vulnerability assessments (Hoq et al., 2021; Raihan & Melon Hossain, 2021), the diversity of land use land cover (LULC) in wetland ecosystems (Alam et al., 2021; Bhattacharjee et al., 2021; Polash et al., 2022), the identification of factors influencing perceptions and impacts of climate change-induced events on livelihoods, along with potential mitigation measures (Baten & Hossain, 2021; Dey et al., 2021; Fahim & Sikder, 2022; Hoq et al., 2022; Tikadar et al., 2022), as well as mapping flash flood susceptibility and conducting hazard assessments (Haque et al., 2021; Islam et al., 2022a, 2022b; Quader et al., 2023), and managing flash flood risks (Abedin et al., 2022). Almost all the recent researchers were to develop flood hazard mapping, detection of climate change impact, change detection analysis, and impact assessment for different projects. An overview of the most recent research and it's focus area is highlighted in Table 1.

Table 1 Recent research at wetland areas in Bangladesh (WoS and Scopus index from 2021 to 2023)
Full size table

Nevertheless, there has not been a comprehensive study that examines the connections among wetlands, climate change, coping strategies, and economic factors, all while integrating advanced satellite remote sensing data. This study fills that gap, aiming to formulate more effective policies for communities in Bangladesh affected by flash floods.

3 Current status of wetland agriculture in Bangladesh

3.1 Wetland statistics at a glance

The wetland region in northeastern Bangladesh spans approximately 1.99 million hectares (ha) and is home to a population of 19.37 million people (Haor Master Plan, 2012). Within this region, there are approximately 373 wetlands spread across seven districts, encompassing an area of approximately 858,460 ha, accounting for roughly 42.93% of the total area of the wetland region (see Fig. 2). Within the districts characterized by wetlands, Sylhet is in the highest position (28.15%), followed by Kishoreganj (26.01%) and Sunamganj (25.47%), respectively.

Fig. 2
figure 2

Source: Adopted and modified from Haor master plan, 2012

District-wise wetland number.

Full size image

3.2 Flash floods and wetlands

Flash flooding is characterized by rapid flooding typically occurring within a timeframe of six hours, often within three hours, following intense rainfall or other triggers. In recent times, there has been a notable increase in the occurrence of flash floods, a prominent natural calamity in the northeastern region of Bangladesh, commonly known as the Wetland Basin (Khan et al., 2012). Figure 3 provides a concise overview of the severe impact of flash floods on the single winter crop in the northeastern wetland region from 2016 to 2020.

Fig. 3
figure 3

Source: Authors’ development from different secondary sources

The historical occurrences of flash floods in the impacted wetland regions of Bangladesh.

Full size image

3.3 An overview of flood time in Bangladesh

Early flash floods start during the month of mid-March to mid-May, which is shown below. Peak flash floods occur during the month of mid-May to mid-July, and sometimes late flash floods occur from mid-August to the end of September (Table 2).

Table 2 Flood time overview in Bangladesh
Full size table

3.4 Statistics of flash flood events in wetland areas (1971–2020)

Table 3 and the pie chart demonstrate that approximately 54% of the total land inundated area falls within the 10–29% range. The analysis of 48 years of data reveals that the majority of inundation falls within moderate categories, with high levels of inundation being fewer common occurrences.

Table 3 Frequency distribution of 48 years (1971–2020) inundation statistics in wetland areas
Full size table

3.5 Land distribution and inundation depth for flash flood-affected areas

From the pie chart, it can be seen that around 56% of land in wetland areas is most prone to flash floods. Still, there might be some other high and medium types of land for cultivation (Fig. 4a). In case of average flood depth of households suffering from the flood at different inundation levels is depicted in Fig. 4b. About 42% of flash floods occur at depth levels greater than 3.6 m, and around 33% lie between 1.8 m and 3.6 m in the wetland areas of Bangladesh.

Fig. 4
figure 4

Source: Authors’ adoption based on Annual Flood Report, 2014

a Land type in wetland areas. b Average flood depth in Sylhet division.

Full size image

3.6 Crop and flash flood calendar at wetland areas in Bangladesh

Rice cultivation in Bangladesh occurs across three distinct seasons: Aus, Aman, and Boro. The Boro season spans from November to May and plays a crucial role in the country’s rice production, which heavily relies on irrigation. The Aus season, occurring from March to August (also referred to as early kharif), involves irrigation for initial crop establishment, followed by early monsoon rainfall for the rest of the season. The Aman season, running from July to December, is another significant period for rice cultivation, primarily dependent on monsoon rainfall (Table 4).

Table 4 Crop calendar for rice production 
Full size table

Understanding the vulnerability of Haor districts to flash floods is crucial for Boro rice farming. This entails having knowledge of the flash flood patterns. In Bangladesh, three types of flash floods are prevalent. They usually occur between March and mid-May, peaking from mid-May to mid-July. Although less frequent, late flash floods from mid-August to mid-October have also been noted in some cases, as shown in Table 5.

Table 5 Flash flood calendar in Bangladesh
Full size table

In contrast with other types of flash floods, an early flash flood is exceptionally destructive as farmers seldom have adequate time to harvest their mature Boro rice crop (please refer to the crop calendar). This crop was constantly threatened with partial loss due to the early occurrence of flash flooding before the maturity stages of Boro rice, posing a significant threat to national food security because the area produced roughly 15–25% of total Boro rice production (BBS, 2017).

4 Climate change scenario in wetland areas of Bangladesh

4.1 Projected statistics of IPCC for temperature and precipitation in Bangladesh

The Intergovernmental Panel on Climate Change (IPCC) projections for temperature and precipitation in Bangladesh indicate concerning changes, with monthly temperatures on the rise. Situation report (2017), in the Fourth Assessment Report of IPCC, predicted the effects of climate change on agriculture, suggesting that a 1–3 °C increase in global mean temperatures by 2100 could lead to decreased cereal productivity in lower latitude regions and increased productivity in higher latitude regions. Confidence levels in these predictions vary, with “low confidence” indicating a 2 out of 10 likelihood of accuracy and “medium confidence” suggesting a 5 out of 10 likelihood. Additionally, the Wetland Basin experiences subsidence due to the collision between the Indian Plate and the Eurasian Plate. The Sylhet sub-basin, in particular, is sinking at a rate of 2.1 cm per year due to downward pressure beneath the Shillong Massif (Johnson & Alam, 1991). However, some studies suggest a lower sinking rate of 2–4 mm/year based on sediment thickness analysis (Goodbred & Kuehl, 2000). Nonetheless, this subsidence is expected to lead to increased annual flooding. The Providing REgional Climates for Impacts Studies (PRECIS) ensemble model forecasts significant increases in rainfall volume and runoff in the wetland region and its upstream catchment during the peak monsoon seasons of the 2020s, 2050s, and 2080s, with consensus among at least 75% of the models. Projections indicate rises of at least 40 mm, 90 mm, and 150 mm in precipitation levels during these periods. Additionally, heightened cloud cover is anticipated over the Meghna catchment, especially over Meghalaya, indicating the possibility of increased rainfall and runoff in the wetland influence region.

4.2 Precipitation level in wetland areas of Bangladesh

A secondary review found that precipitation is the most important variable for causing flash floods. The Bangladesh Meteorological Department (BMD) reports a continuous rise in the frequency of early flash floods. Situated downstream of the world’s highest peak in rainfall (Cherrapunji, Assam), the wetland regions of Bangladesh contend with annual flash floods. A rainfall exceeding 150 mm over a 5–6-day period in the wetland basin and its upstream areas is sufficient to trigger flooding (Biswas et al., 2008). The rainfall events of 2010 and 2017 are noteworthy for their severity. It has been observed that anomalous rainfall patterns occur approximately once in a span of seven years. However, due to global climate change, early flash flood frequency may increase in the future. Additionally, historical data were collected from Sylhet and Srimangal stations of the BMD to show the flash flood-affected month’s precipitation levels for Sylhet and Srimangal stations during Boro rice production.

4.2.1 Precipitation comparison at Sylhet and Srimangal station

Figure 5a and b provide data on the amount of precipitation in the region over a certain period (1970–2016). The graph represents the level of precipitation that shows noticeable fluctuations over the period. In other words, the amount of rainfall varies over time, which is an important aspect to consider when studying the climate and weather patterns in this area.

Fig. 5
figure 5

a and b Historical rainfall patterns of Boro rice production of a Sylhet b Srimangal stations (1970–2016). Source: Authors’ contribution from BARC (2023)

Full size image

Among the months observed in the data, the illustration points out that April and May have the most significant fluctuations in precipitation. The observed findings highlight that between April and May, May experiences the most pronounced fluctuations in precipitation. This means that rainfall levels can change dramatically within a short period in May. Additionally, the trend line shows that in some months, the level of precipitation is above the yearly average rainfall in this wetland area of Bangladesh. This information is essential for understanding the weather patterns in the vulnerable wetland region of Bangladesh, as the end of April to mid-May is Boro rice harvesting time.

4.3 Other significant reasons: temperature variability

In the Boro rice growing season, air temperatures drop below the critical threshold for rice, particularly in the northeastern regions of the country. During the germination stage of Boro rice in wetland areas, there is a possibility of a critical temperature when the minimum temperature goes below 15 °C, which creates spikelet sterility and reduces production. Bangladesh must deal with significant challenges when cultivating Boro rice, especially in the northeastern regions, due to the cold temperatures during the germination and establishment stages. The seedling mortality rate in the northern regions can reach 90% (Rashid & Yasmeen, 2018). Over 2.0 million ha of rice crops in northern and northeastern Bangladesh have been adversely affected by severe cold spells over the past few years. The shorter the time between mid-February and mid-March, the lower the temperature remains above 20 °C, and the greater the chance of spikelet sterility in early-planted Boro rice in wetland areas (Biswas et al., 2008). Despite the potential for flash floods every few years as the Boro rice reaches maturity, farmers in wetland areas view the cultivation of Boro rice as an opportunity (Biswas et al., 2008). Farmers transplant Boro rice seedlings earlier than usual in order to take advantage of the early recession of residual floodwater and reduce the likelihood of flash floods at maturity (Biswas et al., 2008; Rashid & Yasmeen, 2018). In these regions, planting Boro rice at the regular time of year protects the crop from cold injury during the reproductive stage, but it also raises the possibility of flash flooding when the crop reaches maturity (Biswas et al., 2008). According to Biswas et al. (2008), flash flooding typically happens in wetland areas after the second week of April, while Boro rice typically matures by the last week of April. Most farmers have already harvested their Boro rice by this point, but many wetland areas could be impacted by flash flooding as early as the first week of April (Rashid & Yasmeen, 2018).

5 Overall damaged in wetland areas due to flash flood

5.1 Historical damaged scenario in wetland areas

In 2017, a flash flood began on March 28th, impacting six districts of northeastern Bangladesh, including Sylhet, Moulvibazar, Sunamganj, Habiganj, Netrokona, and Kishoreganj. The flood breached embankments in multiple locations, leading to extensive inundation of farmland. It destroyed nearly mature Boro rice crops, covering approximately 219,840 ha of land (Nirapod, 2017). Estimated losses due to flash floods were USD 2 billion and USD 2.3 billion in 1998 and 2004, respectively (World Bank, 2005; ADB, 2005) (Fig. 6).

Fig. 6
figure 6

An overview of overall damage due to flashflood from 1974 to 2017. Note: Except inundation area (Fig. 6a), all other datasets are not available after 2007 (Fig. 6b–h). 2014

Source: Authors’ estimation based Annual Flood Report (2013, 2014), Nirapod (2017)

Full size image

For instance, the flash flood had a severe impact on 518 out of the total 541 unions across 62 subdistricts in six districts of Bangladesh in the year 2017. Approximately one-third of the households in these districts have been adversely affected due to the loss of their Boro rice. Moreover, many residents have also experienced damage to their homes, partially or entirely, and have faced losses in fisheries, domestic animals, and birds (Nirapod, 2017). For numerous decades, the Bangladesh government, in collaboration with various national and international organizations, has been actively combating these flash floods and associated agricultural losses (JNA, 2014; MoEF, 2012; MoEF, 2013; Sikder, 2013). However, the government’s reaction has been swift and substantial. It encompasses provisions such as general relief rice, cash assistance, vulnerable group feeding with an additional special cash grant, employment opportunities, C.I. sheets, and financial aid for housing. What is especially noteworthy is the government’s dedication to transforming wetland areas into thriving economic centers.

5.2 The impact of flash floods on Boro rice yield

Flash floods can significantly impact Boro rice production in the wetland areas of Bangladesh. Boro rice is the major rice variety cultivated during the dry season, and wetland areas are crucial for its cultivation. The given graph shows that the average Boro rice production in the study areas is less than the national average in Bangladesh due to the flash flood occurrence (Fig. 7).

Fig. 7
figure 7

Source: Authors’ computation based on BBS, 2021

Production statistics of Boro rice between Bangladesh and major flash flood-affected districts.

Full size image

6 Vegetation phenology analyses: GIS and remote sensing techniques

6.1 Derivation of vegetation indices

The Sentinel-2A multispectral instrument (MSI) is configured to capture reflectance from various spectral bands, including Blue, Green, Red, and Near-Infrared-1 bands at a 10 m resolution; Red edge 1–3, Near-Infrared (NIR)-2, and Short-Wave Infrared (SWIR) 1 and 2 at a 20 m resolution; and three atmospheric bands (Band 1, Band 9, and Band 10) at a 60 m resolution. A total of fifteen Sentinel-2A MSI images were obtained from April 12th, 2017, 2018, and March 28th for the year 2019, to evaluate the vegetation phenology during the maturity stages of Boro rice. During this stage, selected vegetation indices exhibit the highest sensitivity to changes in the biophysical condition of the observed rice crop (Whitcraft et al., 2015). All satellite data used in this study were sourced from the Copernicus Open Access Hub (https://scihub.copernicus.eu/dhus/#/home), and analysis was performed using ArcGIS 10.8.1®. Our methodology involved employing various vegetation indices as input parameters for developing crop yield prediction models based on vegetation phenology analysis. Hence, normalized difference vegetation index (NDVI), rice growth vegetation index (RGVI), and leaf area index (LAI) were employed to assess the greenness status of the rice crop (González-Betancourt & Mayorga-Ruíz, 2018; Na et al., 2013; Nuarsa et al., 2011). Additionally, normalized difference water index (NDWI) was utilized to examine leaf water content, which is linked to chlorophyll levels and yield variations, while moisture stress index (MSI) was used to evaluate leaf moisture stress before the harvesting stages of Boro rice (Gao, 1996). Vegetation indices were computed using the NIR and Red spectral bands for NDVI and LAI and NIR-SWIR1 for the MSI index. Furthermore, NIR-SWIR1 & 2, along with Red and Blue bands, were employed for NDWI and RGVI indices. Moreover, the linear model of LAI, as proposed by Na et al., (2013), was chosen for its superior accuracy in reflecting NDVI composition compared to exponential and expo-linear models. NDVI, NDWI, RGVI, MSI, and LAI for each image were calculated for 2017, 2018, and 2019 using Eqs. 1, 2, 3, 4, and 5, respectively (Table 6).

Table 6 Overview of the chosen vegetation indices derived from remote sensing spectral bands
Full size table

6.2 Crop phenology analysis based on vegetation indices

Crop phenology analysis based on vegetation indices is an effective approach for monitoring and assessing agricultural dynamics in wetland areas. In Bangladesh, wetlands play a crucial role in supporting the country’s agriculture and biodiversity. Analyzing crop phenology using vegetation indices helps in understanding the timing and duration of crop growth stages, monitoring crop health, and assessing agricultural productivity. The primary vegetation indices used for such evaluations include NDVI, NDWI, RGVI, MSI, and LAI. On average, the values of all five vegetation indices peaked in 2019. The lowest NDWI value, recorded at 0.273 in 2017, was due to image acquisition delays after the flash flood. Conversely, 2018 and 2019 were deemed typical years, with image acquisition aligning with expectations (see Fig. 8).

Fig. 8
figure 8figure 8figure 8

Source: Authors’ computation based on Satellite dataset using ArcGIS 10.8.1 software

Different vegetation indices map for 2017, 2018 and 2019.

Full size image

Generating time series data of the vegetation indices can generate a yield forecasting model for the wetland areas in Bangladesh. This involves extracting the pixel values of the vegetation indices for each image over a specific period. Deriving phenological metrics from the time series data to assess crop phenology is an important step in remote-sensing-based research. Common metrics include the onset of growing season, termination of season, length of growing season, and various growth stages (e.g., emergence, peak growth, senescence). By following proper methods, researcher can conduct a comprehensive crop phenology analysis based on vegetation indices. This information can be valuable for monitoring agricultural productivity, assessing climate change impacts on crops, and supporting decision-making in the agricultural sector. In the later section, probable coping strategies are highlighted based on combinations of economic modeling and satellite remote sensing algorithms for sustainable agricultural production in wetland areas of Bangladesh.

7 Life cycle and proposed coping mechanisms of Boro rice in wetland areas

7.1 Life cycle of Boro rice production

Figure 9 represents the production stages of Boro rice in wetland areas of Bangladesh. Usually, the duration of Boro rice production is longer than any other crop in Bangladesh. It usually takes around 145–160 days for harvest. Land preparation is the initial stage, where the field is prepared by plowing the soil to break it up and create a suitable seedbed for Boro rice planting (SoS), and then the emergence of the early growth of rice plants is shown. Days 60–80 are the tillering stage, while around 120 days after the date of transplanting planting (DAT) is considered the flowering stage. Finally, harvesting/ripening of grain started between 140 and 160 days, which is also flash flood-affected time for wetland areas of Bangladesh (Fig. 9).

Fig. 9
figure 9

Source: Authors’ computation based on Islam et al., (2021b)

The life cycle of Boro rice in wetland areas of Bangladesh.

Full size image

Due to its longer life span, the problem of spikelet sterility is common for Boro rice production in the research area. Typically, farmers in wetland regions opt to transplant Boro rice seedlings ahead of schedule to capitalize on the early recession of residual floodwater, thereby reducing the vulnerability to flash flooding during the maturity stage. However, this early transplantation exposes Boro rice to lower temperatures during the reproductive phase in February, leading to increased sterility rates (Biswas et al., 2008; Rashid & Yasmeen, 2018). Farmers commonly choose to grow BRRI dhan29, a high-yielding rice variety with a long growth period of 160 days. They usually plant their seeds in seedbeds between late October and early November, and the crop matures by mid-April. While BRRI dhan29 is generally resilient to cold injury during the reproductive stage because of its extended growth period, it remains vulnerable to flash flooding at maturity (Table 7).

Table 7 Sterility triggered by planting timing and the susceptibility of Boro rice to flash floods
Full size table

Additionally, there is a risk of spikelet sterility when planting a short-duration variety such as BRRI dhan28 on November 1st. Recently, Bangladesh Rice Research Institute (BRRI) has identified a moderate level of cold tolerance in a previously developed Boro variety, BRRI dhan69, during both the seedling and reproductive stages. However, BRRI dhan69 has a slightly longer growth duration compared to BRRI dhan28, spanning approximately one week more. Employing BRRI dhan69, recognized for its cold resilience during the reproductive phase, could alleviate sterility problems linked to early transplantation. So, people in wetland areas face twofold problems with Boro rice cultivation. In the next section, some proposed alternative coping mechanisms have been discussed.

7.2 Alternative coping mechanisms for sustainable agricultural production

7.2.1 Land suitability assessment

In the wetland areas prone to flash floods, there are several alternative coping possibilities that farmers can consider. These crops are adapted to the wetland environment and can provide sustainable income and food security. The land resources information system of Bangladesh Agricultural Research Council (BARC) provided agro-edaphic and agro-climatic data to evaluate land suitability for specific crop cultivation. The evaluation analyzed eleven agro-edaphic parameters, comprising soil permeability, adequate soil depth, available soil moisture, nutrient status, pH, soil salinity, soil consistency, drainage, depth of inundation, flood hazards, and slope. Additionally, agro-climatic factors, including the duration of the kharif growing season, pre-kharif transition period, thermal zone, and extreme temperatures, were considered to assess their impact on crop growth concerning crop phenology and photosynthesis. Crop suitability based on agro-edaphic and agro-climatic factors was analyzed separately, considering the limitations imposed by each factor concerning crop requirements. The current study considered the most vulnerable subdistricts from the wetland region of Bangladesh, including Tahirpur, Gowainghat, and Kulaura. From the findings of the Tahirpur subdistrict, it is seen that chili can be produced instead of Boro rice as more than 20% of the land is in the suitable category, while mustard is in the moderately suitable category (Fig. 10).

Fig. 10
figure 10

Source: Authors’ computation based on BARC (2023)

a–c Potentially suitable land for major crops at a Tahirpur; b Gowainghat and c Kulaura subdistrict in Bangladesh.

Full size image

On the other hand, in the Gowainghat subdistrict, mustard has much more opportunity to grow, as 30% of the land is shown to be suitable. Interestingly, Kulaura is best positioned to practice alternative crops to secure their production by changing cropping patterns with mungbean, mustard, and jute. Based on the suitability assessment findings, some alternative crops can be introduced to mitigate the risk of flash floods in Bangladesh. Hence, it is crucial to identify and delineate suitable areas for growing particular crops to harvest maximum potential yield and combat the adverse climatic impact.

7.2.2 Flash flood vulnerability mapping

Vulnerability mapping for flash floods and livelihood vulnerability assessments are frequently regarded as necessary components of risk mitigation strategies. To secure farmers’ livelihoods, thorough and advanced policy design is essential as the occurrence and severity of extreme events escalate, and vital crops suffer damage due to global climate change. As a result, employing strategies like flash flood risk assessment, prevalence mapping, monitoring, and vulnerability analysis becomes an essential proactive management measure, particularly for low-lying regions, and has significant implications. Recently, flash flood vulnerability mapping has gained prominence as a pertinent and significant approach for identifying flash flood-prone zones and informing future decision-making (Islam et al., 2022a, 2022b). The main emphasis is on identifying and sharing the most appropriate and efficient adaptation strategies derived from vulnerability assessments with all vulnerable communities. Thus, employing flash flood risk mapping and conducting need assessments through field observations could prove pivotal in reducing livelihood vulnerability (see Fig. 11).

Fig. 11
figure 11

Flash flood vulnerability map. Source: Authors’ computation based on Islam et al., (2022b)

Full size image

7.2.3 Yield forecasting for decision support system

Forecasting crop yields prior to harvest is crucial for managing natural disasters, ensuring food and nutritional security, predicting market prices of agricultural commodities, as well as alleviating poverty. Moreover, early prediction of significant crop yields enables policymakers to gauge the necessary amounts of imports and exports. Extreme natural phenomena caused by climate change, including flash floods, regularly harm almost mature rice in wetland areas of Bangladesh as well as other developing nations. Therefore, early crop yield prediction before harvest can substantially aid in reducing losses and achieving desired yields and earnings in this area. Moreover, given the adverse impacts of environmental changes, it is essential to explore yield prediction models during the mature stages of crop production (Islam et al., 2021a, 2021b). To understand the spatial distribution of projected rice yields in the study area, each map underwent segmentation into four categories using the manual brake reclassification method within the ArcGIS environment. The results of this reclassification are consistent with ground reference data and can aid in mapping Boro rice yields before and after flash flooding during the monsoon season. Additionally, the analysis revealed that the highest estimated yield, noted in both simple regression analysis (4.25 MT/ha) and multiple regression analysis (4.23 MT/ha), occurred in 2018, with yields varying from 0.11 to 4.25 MT/ha. The predictive yield models for 2018 and 2019 demonstrated improved performance compared to the model for 2017, a year characterized by flash flooding (refer to Fig. 12a–d).

Fig. 12
figure 12

Source: Authors’ computation based on Islam et al., (2021b)

Illustrates the projected yield rates for Boro rice using composite vegetation indices. Panel a shows the data for 2017, while panel b depicts 2018, and panel c represents 2019. Panel d presents the average data from 2017 to 2019.

Full size image

Hence, the forecasting models derived from satellite data and historical yield information can estimate rice yields during the rice crop’s maturity stages, allowing for an assessment of potential harvest opportunities before flash floods occur. Looking ahead, these models can be applied to evaluate the risk associated with rice yield production, especially in the context of extreme climate events brought about by climate change, thus aiding in supporting farmers’ livelihoods.

7.2.4 Crop insurance: an effective and fruitful risk mitigation strategy

In South Asia, people’s livelihoods are inextricably linked to agricultural productivity, making them weather-dependent. Agriculture has relied on the annual monsoon rains for millennia. Nonetheless, because of climate change, the timing and intensity of this yearly growth are becoming more irregular, putting farmers’ livelihoods at risk. Crop insurance is attracting much attention from the international development community to reduce these risks. Furthermore, it is essential to implement site-specific economic strategies to optimize agricultural subsidies based on the severity of the damage, aiming to enhance the sustainable livelihoods of farmers impacted by the flash flood in wetland areas of Bangladesh. Establishing a cost-effective insurance scheme in Bangladesh is crucial to aid farmers in managing risks and fostering the establishment of long-term insurance markets specifically designed for small-scale farmers. Furthermore, it is important for all crop insurance policies to encompass risks related to weather conditions and, in certain instances, natural calamities. Therefore, developing a damage-based crop insurance model that considers constraints like data limitations, knowledge gaps, farmer awareness, costs, and experience could serve as an effective resilience strategy in wetland areas of Bangladesh. This approach has significant potential to help farmers mitigate risks associated with flash floods (Islam et al., 2021a). Figure 13 depicts a graphical representation of a damage-based crop insurance model.

Fig. 13
figure 13

Source: Authors’ own development

Damage-based crop insurance model against flash floods.

Full size image

8 Existing problems and possible ways to tackle the difficulties

Wetland farmers in Bangladesh face several challenges and problems that impact their livelihoods and agricultural practices. Wetland areas are highly vulnerable to flash floods and waterlogging, damaging crops, destroying infrastructure, and disrupting farming activities. Sudden inundation of fields can result in crop losses, soil erosion, and the destruction of irrigation systems. Many wetland areas in Bangladesh lack adequate drainage systems to manage excess water during heavy rains or floods. Insufficient drainage infrastructure leads to prolonged waterlogging, making it difficult for farmers to cultivate crops or rear livestock. Wetland farmers often need help accessing formal credit and financial services. Due to the high-risk nature of farming in wetlands, financial institutions may hesitate to provide farmers with loans or credit facilities. Access to finance is needed to ensure investment in agricultural inputs, technology, and infrastructure. Wetland farmers often face challenges accessing high-quality seeds, fertilizers, and other agricultural inputs. Availability and affordability of inputs, such as improved seeds and nutrient-rich fertilizers, are crucial for enhancing crop productivity and yields. Many wetland farmers need access to information, training, and technical support to adopt modern farming techniques and climate-resilient practices. This includes knowledge of appropriate crop varieties, water management techniques, pest and disease control, and sustainable farming practices. Wetland farmers often face difficulties in accessing markets and obtaining fair prices for their produce. Limited transportation infrastructure, inadequate storage facilities, and lack of market linkages contribute to challenges in marketing and selling their agricultural products. Climate change exacerbates the challenges faced by wetland farmers in Bangladesh. Erratic rainfall patterns, increased temperature, and rising sea levels contribute to higher vulnerability to floods, and changes in the availability of water resources. These climate change impacts further threaten agricultural productivity and livelihoods in wetland areas.

However, addressing these challenges requires comprehensive measures and support from various stakeholders, including the government, non-governmental organizations (NGOs), and development organizations. This includes climate-resilient agricultural practices, ecosystem-based adaptation, improved water management, capacity building, and knowledge transfer. Economic analysis can identify and promote sustainable farming practices that enhance resilience to climate change, such as the introduction of crop insurance model, integrated water management, crop diversification, and agroforestry. Additionally, satellite data can provide insights into the spatial distribution of suitable areas for these practices and monitor their effectiveness. Satellite data can also help identify priority areas for conservation and restoration efforts. Furthermore, satellite technology can assist in monitoring water availability, including changes in precipitation and water storage, and support the development of water management strategies. Moreover, preserving and restoring wetland ecosystems can provide natural buffers against the impacts of climate change. Economic assessment techniques can evaluate the ecosystem services provided by wetlands, including water purification, flood control, and carbon storage. Economic analysis can also guide the design of efficient and equitable water allocation mechanisms in wetland agricultural systems.

While the research offers valuable perspectives on the interplay among wetlands, climate change, and adaptive strategies concerning Boro rice cultivation in Bangladesh, it is essential to acknowledge certain constraints: First, this study did not follow any systematic review methods. Second, the climate change impact scenarios present general findings rather than econometric modeling-based empirical estimations. Third, the assessments of crop phenology and coping mechanisms rely on the author’s previous research findings to support the results. Fourth, while the study mentions incorporating economic considerations, it may benefit from a more detailed analysis of the economic implications of different adaptation strategies, including costs, benefits, and potential trade-offs. Finally, while the study identifies adaptation strategies, it needs to investigate the potential challenges and barriers to implementing them at a policy or field level. Therefore, a more comprehensive study from similar ground reference points (field survey) would yield more robust results. Future research should consider all these crucial dimensions and more advanced econometric and spatial modeling to examine the relationships among climate change, wetland agriculture, and copping mechanism nexus simultaneously.

It is important to note that the solutions proposed here are incomplete and should be tailored to the specific context of each wetland agricultural system. Furthermore, effective implementation requires interdisciplinary collaboration among economists, ecologists, agricultural scientists, policymakers, and local communities. By combining economic principles with satellite technology, we can deepen our comprehension of the interconnection among wetland farming, climate change, and adaptation strategies, thereby advancing the development of sustainable and resilient agricultural systems amid shifting climatic conditions.

9 Conclusion and policy implications

This integrative review explores the intricate interplay between climate change, coping mechanisms, and Boro rice cultivation in the wetland areas of Bangladesh. Utilizing an integrated approach encompassing economics and satellite technologies, this study aims to dissect the multifaceted dynamics in this critical agricultural landscape. In this investigation, the study utilized optical remote sensing imagery to examine the effects of flash floods on Boro rice production by analyzing inundation maps in the districts of Bangladesh susceptible to flash floods. The present research provided a basic overview of flash flood-affected wetland areas, considering climate change aspects (monthly maximum, minimum temperature, and precipitation) and its effect on Boro rice production. Although secondary data from the IPCC indicated an alarming outlook for the monthly temperature and precipitation change in Bangladesh, the trend indicated an escalating rate. If these anticipated outcomes come to pass, a shocking situation awaits the wetlands in Bangladesh, particularly in areas affected by flash floods. The study showcased the efficacy of employing Sentinel-2A-derived time series data for mapping submerged wetland regions in 2017 and tried to compare them with normal-year vegetation indices through NDWI, NDVI, and soil-adjusted vegetation index (SAVI) analyses. Our results showed that the value of NDWI before a flash flood is much better than after a flash. The SAVI and NDVI analyses for the flash flood-affected wetland areas also support this result. Despite the proposed methods’ accuracy, the study strongly recommends thoroughly investigating their evaluation before applying it to different regions.

Consequently, some probable coping mechanisms, including the evaluation of land suitability assessment, the demarcation of flash flood-vulnerable areas, the development of a yield forecasting model, and the introduction of a crop insurance model, were also discussed in this paper. The land suitability assessment shows a promising indication that it should be implemented without delay to ensure sustainable crop production in the flash flood-affected region of Bangladesh. The results from the vulnerability mapping indicate that a detailed and reliable flash flood vulnerability map, achieving an accuracy level exceeding 75%, can be created using readily available criteria known to trigger flash floods.

Furthermore, the yield prediction models, developed from satellite data and historical yield information, offer a means to forecast rice yields during maturity stages, allowing for the assessment of harvesting prospects ahead of probable flash flood events. Finally, the study introduced a straightforward but inclusive, time-demanding, and appropriate microinsurance model based on flash flood-induced damaged classifiers (high, moderate, and marginal) and their coverage preferences. Therefore, the concerned authority should investigate the feasibility of implementing a community-centered microinsurance model in more flood-prone regions. Hence, it is imperative to implement decisive and impactful measures promptly.