The research was conducted in Khulna District, located in the coastal region of Bangladesh. The study focuses on six sites within Batiaghata, a sub-district of Khulna. Batiaghata is divided into two parts, western and eastern, by the Kajibacha River. It is bounded by Rupsa, Rampal and Fakirhat upazilas to the east; Dumuria and Paikgaccha upazilas to the west; Kotowali and Sonadanga thanas and Dumuria upazila to the north; and Dakope, Rampal and Paikgachha upazilas to the south. The area of Batiaghata covers 235.32 square kilometers and is situated between latitudes 22° 34' and 22° 46' north and longitudes 89° 24' and 89° 37' east.

According to the 2022 Bangladeshi census, the population of Batiaghata was 171,752, with a population density of 730 per square kilometre, comprising 86,685 males and 85,067 females. The main rivers in Batiaghata upazila are the Kazibachha, Shoilmari, Pasur and Vadra. The annual average maximum and minimum rainfall totals in the region are 172.60 mm and 152.40 mm respectively, with average maximum and minimum temperatures of 31.7°C and 22.3°C. Figure 1 and Table 1 present the selected study sites.

Electrical conductivity (EC) data from six sites in Batiaghata were collected from the Soil Research Development Institute (SRDI), at Khulna in Bangladesh Table 2. Satellite Landsat 8–9 OLI data for the period from January 2023 to December 2023 were downloaded from the USGS Earth Explorer data portal Table 2. In addition, severe cyclonic storm data were collected from the India Meteorological Department (IMD).

The methodology is subdivided into three parts:

(i) Data collection; (ii) Processing and mapping; and (iii) Statistical analysis

Data collection : For the study, climatic driver data (rainfall and temperature) were collected from BMD, cyclone data from IMD, and EC data from SRDI. The data were then processed using Excel software to analyse seasonal variation and trends through regression equations. Subsequently, freely accessible remote sensing data were collected from the Landsat program (https://earthexplorer.usgs.gov/), which involves a series of Earth-observing satellite missions jointly managed by NASA and the US Geological Survey (USGS).

Processing and mapping : Monthly climatic data were obtained from the daily data. From the monthly data, seasonal mean values were computed for the four seasons, pre-monsoon (March-May), monsoon (June-September), post-monsoon (October-November) and winter (December-February), in an Excel template. Additionally, the average EC value was evaluated for the Khulna district from the six sites. The influence of climatic conditions on soil salinity was evaluated using a bar graph. Satellite data extraction, raster calculation, and the creation of cartographic materials were then conducted in the geographic information system (GIS) for the mapping of indices. The seasonal field data of soil salinity and satellite indices were compared to determine the relationship between them based on maximum and minimum values.

Statistical analysis : The coefficient of determination (R 2 ) was obtained through diagrams, while the significance of the R 2 was assessed using a t-test.

Using remote sensing data, especially Landsat imagery, NDSI demonstrated encouraging results for the assessment of soil salinity. Additionally, NDSI frequently shows a high positive association with electrical conductivity (EC) recorded in the field. Since NDSI performs well in detecting soil salinity in areas with minimal vegetation, and as the study area is a coastal region with low vegetation, it is effective in representing soil salinity in the local context.

NDSI indices were derived by using the formula shown in Table 3 for Landsat 8-9 OLI.

R (Red) = Band 4 (0.64 – 0.67 µm)

NIR (Near Infrared) = Band 5 (0.85 – 0.88 µm)

Soil salinity across the six sites in the coastal district of Khulna exhibits both spatial and seasonal variability. The highest salinity levels were recorded during the pre-monsoon season, while the lowest occurred in the post-monsoon season Figure 2. As elevated salinity during the pre-monsoon period adversely affects crop production, the use of rainwater irrigation and the adoption of salt-tolerant crop varieties are recommended as effective mitigation strategies to enhance agricultural productivity in the region.

In 2021, the highest soil salinity during the pre-monsoon season was observed at Fultala_2 (14.5 dS/m), followed by Fultala_1 and Kismat Fultala_2, while the lowest maximum salinity was recorded at Krishnanagar_1 (11.6 dS/m). These peak values reflect the seasonal salinity buildup prior to the monsoon. In contrast, the lowest soil salinity values occurred during the monsoon and post-monsoon seasons, with the highest minimum level recorded at Krishnanagar_1 (1.5 dS/m in 2020), and the lowest at Fultala_1 (1.1 dS/m in 2021) (see Table 4). These findings highlight a clear seasonal pattern: pre-monsoon salinity levels remain higher than those of any other season, emphasising the strong influence of climatic factors on soil salinity dynamics across all the study sites.

Figure 3 illustrates the seasonal trend of soil salinity for the Khulna district, aggregated from individual site measurements. The data reveal a consistent pattern, with peak salinity observed during the pre-monsoon season and the lowest levels during the post-monsoon season, mirroring the trends identified at individual sites. This consistency indicates a uniform seasonal salinity pattern across the study area.

Figure 4 demonstrates that electrical conductivity (EC) values consistently peak during the pre-monsoon season across all sites and years. The levels decline from pre-monsoon to monsoon periods, followed by a gradual increase through the post-monsoon and winter seasons, forming a cyclic pattern influenced by climatic factors. The maximum EC values generally demonstrated an upward trend, with the minimum values also indicating a gradual baseline increase in soil salinity, which may pose long-term risks to agricultural sustainability. Increases in the maximum and minimum EC values influence the crop tolerance range, consequently hindering crop productivity.

AlthThough the highest EC values were observed during the pre-monsoon season Figure 2, a decreasing trend is evident across all sites during the study period Table 5. In contrast, increasing trends are observed during the monsoon and post-monsoon seasons. The winter season exhibits both increasing and decreasing trends, depending on the site. The highest rate of increase in soil salinity was recorded at Kismat Fultala_2 during the post-monsoon season (0.966 dSm-1/year), which is statistically significant at the 90% level of significance, while the lowest was also at Kismat Fultala_2 in the winter period (0.09 dSm-1/year). Conversely, the most significant decrease occurred at Fultala_2 in the pre-monsoon season (-0.6967 dSm-1/year), while the smallest decrease was at Krishnanagar_2 during the winter season (-0.0993 dSm-1/year). These trends highlight seasonal and spatial variability in soil salinity dynamics across the study area.

Rainfall reduces soil salinity by diluting salt concentrations through leaching, while high evaporation increases it by removing soil moisture and leaving salts behind. In Bangladesh, rainfall is highest during the monsoon season and lowest in winter. Figure 5 (a–d) illustrates the relationship between soil salinity and climatic parameters, namely rainfall and temperature, for the Khulna region. Excessive rainfall contributes to reduced soil salinity in the monsoon season. In contrast, the lower rainfall during the post-monsoon and winter seasons leads to increased salinity in winter, peaking in the pre-monsoon season, due to continued moisture deficit. Elevated average temperatures during the pre-monsoon season further intensify evaporation, exacerbating salinity levels. Conversely, lower average temperatures correlate with reduced salinity. These findings suggest a strong inverse relationship between rainfall and soil salinity, and a positive correlation between temperature and salinity, indicating the significant influence of climatic factors on soil salinity dynamics in coastal Bangladesh.

Khulna faced a tropical cyclone from 30 November to 6 December, 2021, namely JAWAD Table 6. Storm surges, commonly associated with such events, transport saline water inland, leading to increased soil salinity through salt deposition. The impact of tropical cyclones on soil salinity is clearly illustrated in Figure 6. Following the cyclone event referred to above, soil salinity exhibited an exceptional increase in December 2021 compared to the same month in previous and subsequent years.

This indicates a sustained influence of storm surge events on long-term soil salinisation in coastal areas. Specifically, EC levels sharply increased in December 2021, suggesting that a notable event had occurred during this transition period compared to other years. This event may have been associated with the storm surge of tropical cyclone JAWAD.

Standard deviation is a fundamental statistical metric that quantifies the degree of dispersion or variability of data points relative to the mean of a dataset. A higher standard deviation indicates a broader spread of values, suggesting that the data points are more widely distributed. In contrast, a lower standard deviation signifies that the data are more tightly clustered around the mean. The values of NDSI, which are computed using spectral reflectance data, normally fall between -1 and +1. Higher salinity is correlated with positive NDSI levels, and lower levels with negative values. Positive and negative values are produced because they arehigher and lower than the reference value respectively. In this study, NDSI values were calibrated to match field-measured electrical conductivity (EC) data to enhance the accuracy of salinity detection. Satellite-derived NDSI values were processed and visualised using ArcGIS, as shown in Figure 7 (a–d).

The highest NDSI values were observed during the pre-monsoon season, aligning with EC data obtained from SRDI, Khulna (Table 7 and Figure 8). A clear seasonal cyclic variation is evident in Figure 7 (a–d) and Figure 8, showing strong correlation with ground data, except for some inconsistency during the monsoon season. This ambiguity may have occurred due to the comparisons being made between the direct field observation values and the maximum and minimum values of the satellite data, which incorporate high spatial resolution.