Prerpints.org logo
Preprint
Article

Assessing Land Use and Land Cover Dynamics of Irrawaddy Delta: Remote Sensing Analysis and Future Projections

Altmetrics

Downloads

188

Views

185

Comments

0

supplementary.docx (14.24KB )

Submitted:

03 July 2024

Posted:

04 July 2024

You are already at the latest version

Alerts
Abstract
Significant changes in Land Use and Land Cover (LULC) have widespread implications for the environment, economy, and society; influencing future sustainability and development of a region. This study aimed to assess LULC changes for a 30 year period (1990 – 2020) and project future LULC from 2030 to 2100 for the Irrawaddy Delta using remote sensing and simulations with the artificial neural networks-cellular automata method. The findings showed significant LULC changes in the Delta, particularly for the mangrove forests and cropland (rice paddies). Mangrove coverage was 1,471 km² in 1990 but decreased to 1,282 km² in 2020, and simulations predict a further reduction to 1,277 km² in 2100. Cropland areas increased from 10,550 km² to 10,618 km² in 2020, with simulations projecting a slight decrease to 10,586 km² by 2100. Human activities and cyclonic events, especially in the southern part of the region, have led to barren land and impervious areas replacing dense mangrove and forested areas. These changes threaten the ecological integrity of the Delta and impact local livelihoods and biodiversity. The study underscores the need for sustainable land management practices and policies to mitigate the adverse effects of LULC changes and ensure the resilience of the Irrawaddy Delta against ongoing environmental challenges.
Keywords: 
Subject: Environmental and Earth Sciences  -   Remote Sensing

1. Introduction

Approximately half a billion individuals worldwide live in or near delta plains, many in large coastal cities [1]. The 20th century saw significant impacts on deltas from population growth, economic expansion, and catchment development [2]. Over 10 million people are affected annually by storm surge flooding, mostly in low-lying river deltas in South Asia. In particular, the Irrawaddy Delta, also known as the Ayeyarwady Delta, faces increasing risks of coastal flooding due to land subsidence outpacing sea level rise [3]. This subsidence exacerbates the effects of rising sea level, making the region more susceptible to flooding and coastal inundation. Additionally, the Delta is experiencing a significant loss of biodiversity, which threatens the ecological balance and the services that these ecosystems provide [4,5]. The heightened vulnerability to storm surges poses a further greater threat to natural habitats and human settlements, as severe storms can cause extensive damage to infrastructure and agricultural lands [6,7].
Shoreline erosion is another pressing issue, with approximately 60% of the shoreline undergoing chronic erosion along the western flank of the Irrawaddy Delta [8], leading to persistent loss of land and displacement of communities. Saltwater intrusion into freshwater systems and agricultural lands is compromising water quality and soil fertility, making it increasingly challenging for local farmers to maintain crop yields. Degradation of wetlands, which serve as crucial buffers against floods and support a diverse range of species, is further diminishing the region's natural defenses [9]. Intensive agriculture practices, driven by the need to support a high population density, are putting additional pressure on the Delta's resources. Moreover, the region's high population density increases the demand for land and other resources, contributing to the overexploitation and even the destruction of natural habitats [10]. Ineffective governance and limited adaptation capacity exacerbate these challenges.
Rivers play a crucial role in transferring water and sediment from its watershed to the coast, profoundly influencing delta formation and its evolution. Human activities and land-use changes have significantly altered streamflow and sediment transport, reducing sediment supply to coastal regions [11,12,13,14]. This is particularly relevant for Myanmar, where the country's climate has experienced significant shifts in recent decades, resulting in frequent natural disasters such as coastal flooding, droughts, heat waves, and cyclones [15,16]. Because of its low elevation, the Irrawaddy Delta is particularly vulnerable to flooding from tropical cyclones and storms. In Myanmar, the Irrawaddy River Delta, one of the largest deltas in the world, covers 60% of the country's territory and is home to about 90% of its population. In 2008, Cyclone Nargis inundated 40% of the delta, killed more than 130,000 people, and caused the shoreline to retreat for an average of 47 kilometers [6]. Despite ranking second globally in the Global Climate Risk Index between 1995 and 2014 [13], there have been few comprehensive assessments of how climate change and human activities impact Myanmar's river and deltaic systems [17,18]. The delta has experienced significant ecological disturbances from population growth and developmental activities such as mechanized farming, dam construction, deforestation, and mining [19]. Sirisena et al. [11] projected that the construction of planned upstream reservoirs could decrease monsoon streamflow by 6-7% and sediment load during the same period by 9-11%. Recent developments, including large-scale reservoir construction and extensive in-channel sand mining, have disrupted river flow and sediment dynamics; leading to delta subsidence and reduced resilience [20,21]. To effectively develop a sustainable management plan for the Irrawaddy River Basin and its Delta, it is essential to understand and assess the impacts of Land Use and Land Cover (LULC) changes. Long-term data analysis and prediction could help in implementing measures for flood and drought mitigation and adaptation, designing sustainable river structures for training the channels, and maintaining inland water bodies [21].
Over the past half a century, the Irrawaddy Delta has experienced significant LULC changes driven primarily by agricultural expansion, urbanization, and the impacts of climate change [21,22]. As a result, the mangroves in the Delta have been converted into croplands, predominantly rice paddies. This transformation has been documented by Renaud et al. [23] and; Win et al. [7]. Furthermore, there was an annual deforestation rate of 1.7% recorded from 2010 to 2015 [24]. While Myanmar’s mangrove areas only make up about 6% of the total in Southeast Asia [25], they are crucial to the country’s ecology and economy. Key factors contributing to the degradation of mangrove forests in Myanmar include land allocation for human settlements, unplanned expansion of rice paddies, overuse of mangrove resources for fuel and timber, and the migration of people from inland areas to coastal regions, prompted by population growth and political instability [26,27]. The conversion of these natural habitats has not only altered the landscape but also had adverse effects on the environment and livelihoods of local populace who heavily depends on traditional farming and fishing for their subsistence . The expansion of agricultural activities, especially rice cultivation, has resulted in the large-scale clearing of mangrove forests, which are essential for maintaining ecological balance and protecting coastal areas from erosion. Simultaneously, urbanization has led to the development of infrastructure such as roads, industrial zones, and settlements, further encroaching on both agricultural and natural lands [28]. Climate change has exacerbated these trends, with rising sea levels and increased frequency of cyclones causing additional stress on these low-lying ecosystems [29]. Furthermore, the sustainability of agriculture is threatened by soil degradation and salinization, which undermine long-term productivity and food security. These multifaceted impacts highlight the urgent need for developing integrated land management strategies that balance development with conservation efforts to ensure the resilience and sustainability of the Irrawaddy Delta's ecosystems and the communities they support.
To address the complex LULC changes in the Irrawaddy Delta, advanced modeling techniques such as Cellular Automata-Artificial Neural Networks (CA-ANN) offer a promising solution for predicting and managing these transitions. CA-ANN integrates the spatial dynamics of cellular automata (CA), which simulate the spread and interaction of LULC types based on predefined transition rules, with the pattern recognition capabilities of artificial neural networks (ANN) [30,31,32,33]. ANN can analyze complex relationships between influencing factors like socio-economic, climatic, and environmental variables, offering insights into the underlying patterns driving LULC changes. CA-ANN is often considered superior to other methods due to its ability to effectively represent nonlinear spatially stochastic LULC change processes [34]. It can successfully handle noisy data and outliers, provide considerable classification accuracy, and streamline the LULC classification process by focusing on essential factors [35]. Furthermore, the CA-ANN model can generate rich patterns and effectively represent nonlinear spatially stochastic LULC change processes [30,34,36]. Several studies have demonstrated the effectiveness of CA-ANN in predicting LULC changes. For instance, Asadi et al. [37] found that while the CA-Markov model was more accurate in predicting larger areas, ANN outperformed in predicting smaller areas. Xing et al. [38] enhanced the model by integrating CA with deep learning, resulting in a 9.3% to 11.67% increase in prediction accuracy. Omrani et al. [39] extended the model to include the multi-label concept, which improved the goodness-of-fit calibration values. Ahmad et al. [30,31] demonstrated that CA-ANN and CA-Weight of Evidence performed well in predicting the future land use and land cover (LULC) of the Sundarbans Delta. Moskolai et al. [40] compared CA-Markov with deep learning (DL) algorithms and found that the latter, particularly convolutional long short-term memory (ConvLSTM) networks, were better suited for LULC predictions. These studies collectively highlight the potential of CA-ANN in refining and improving the accuracy of LULC predictions, especially for deltaic regions that are vulnerable to a myriad of environmental stressors.
The primary objective of this study was to assess the long-term evolution of LULC patterns in the Irrawaddy Delta, focusing specifically on mangrove vegetation and cropland expansion in the region's most productive, yet highly vulnerable coastal zone. The study analyzed classified images to understand how LULC trends impact the delta's response to sea level rise and ever-reducing terrigenous sediment supply from rivers. Using a CAA-ANN, we simulated the future LULC of the delta, considering influential factors such as Land surface temperature, vegetation index, and Representative Concentration Pathways (RCPs) scenarios.

2. Materials and Methods

2.1. Study Area

The study was conducted for the lower half of the Irrawaddy Delta, which is located in the Irrawaddy Division of Myanmar. The Delta spans approximately 60% of Myanmar's land area and supports over 90% of the population, who heavily depend on the river and its resources for their livelihoods [8,41]. The Irrawaddy Delta, one of Myanmar's most densely populated regions with about 15 million residents in its 35,000 km² area, includes 89 Key Biodiversity Areas [42,43]. This low-lying area extends from Myan Aung to the Bay of Bengal and the Andaman Sea, covering 290 kilometers at the mouth of the Ayeyarwady River. The modern Delta is thought to have originated approximately 7000-8000 years ago, which is around the same time as the formation of other major deltas in Southeast Asia [21]. This wedge-shaped Delta extends into the Gulf of Martaban and covers approximately 20,571 km2 of flat, low-lying land with fertile soil and includes five major and many smaller distributaries [44,45]. It is characterized by mud and silt dominance, with substantial tidal influence; averaging a tidal range of 4.2 m [46]. Tidal influence extends deep inland, reaching as far as 300 km upstream and causing saline water to penetrate 100 km upstream. Along the coast, the maximum spring tidal range increases eastward in the Gulf of Martaban, peaking at 6.4 meters at Elephant Point. During simultaneous ebb tides and river flooding, the Bassein estuary, located within the western side of the delta, experiences strong tidal currents reaching speeds of up to 3 meters per second [47]. The Irrawaddy River basin has a tropical monsoon climate, with most of the rainfall occurring during the hot and humid months of the southwest monsoon - from May to October. In contrast, the northwest monsoon from December to March is relatively cool and dry, with winds blowing in the opposite direction from the northeast. On average, the Delta receives between 2000 and 3000 mm of annual rainfall [48]. The melting of snow and glaciers in mountainous northern Myanmar – an eastward extension of the Himalayas- during the summer also contributes to the variability of water flow in the Ayeyarwady River. Heavy rainfall in mountains coupled with spring snowmelt can generate large volumes of water that can cause widespread downstream overbank flooding [42]. The most recent estimate is that the annual sediment load brought in by the Irrawaddy River is ~ 364±60×106 tons, and the average annual discharge varies from 422×109 to 440×109 cubic meters [49,50]. The total discharge of the Irrawaddy and Chindwin Rivers, the major tributaries of Irrawaddy and originating at the Indo-Myanmar border, varies greatly, ranging from 1500 to 30,000 cubic meters/second, and that of the head of the Delta varies from 2000 to 33,000 cubic meters/second [51]. The data collected by Robinson et al. [52] and analyzed by Furuichi et al. [41] also estimated the water discharge and suspended sediment load upstream of the Delta head at 379 ± 47×109 cubic meters per year and 325 ± 57×106 tons per year, respectively.
Preprints 111177 g001
Figure 1. Study area: Irrawaddy River Delta. The red-colored polygon highlights the selected area, which is located in the southern half of the Delta. The adjoining Gulf of Martaban is highly turbid with heavy discharge from the Irrawaddy River and its distributaries during springtime.
Figure 1. Study area: Irrawaddy River Delta. The red-colored polygon highlights the selected area, which is located in the southern half of the Delta. The adjoining Gulf of Martaban is highly turbid with heavy discharge from the Irrawaddy River and its distributaries during springtime.
Preprints 111177 g001

2.2. Data Collection

This study used Landsat sensor images from the Landsat 5 and Landsat 8 missions at ten-year intervals from 2000 to 2020 (Table 1). These images were used to extract the Land Surface Temperature (LST), Normalized Difference Vegetation Index (NDVI), and Normalized Difference Water Index (NDWI). Additionally, elevation and slope data were derived from NASA's Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) at a 30-meter resolution, while forest loss data were sourced from the Hansen dataset [53] on the Google Earth Engine (GEE) platform. ‘Forest Loss’ is defined as a stand-replacement disturbance, or a change from a forest to a non-forest state, during the period 2000–2023. Vector data from OpenStreetMap were also downloaded to support the LULC analysis and prediction. The GEE platform, renowned for its robust cloud-based geospatial analysis capabilities, enabled the efficient processing of large volumes of satellite imagery, offering comprehensive and timely insights for informed decision-making. LULC data were obtained from Zhang et al. [54], integrating Landsat time series with high-quality training data from the Global Spatial Temporal Spectra Library on GEE.

2.3. Spatial Indices and Land Surface Temperature Calculation

The normalized difference vegetation index (NDVI) is a remote sensing index used to assess the greenness of vegetation [55]. It is calculated from the reflectance values of visible and near-infrared light by satellite or aircraft sensors. The NDVI is calculated using the following Equation (1)
N D V I = ( N I R − R e d )   /   ( N I R + R e d )
Where NIR is the reflectance of near-infrared light and Red is the reflectance of red light. NDVI values range from -1 to 1, with higher values indicating more green vegetation and negative values indicating water bodies. NDVI is commonly used to monitor and assess vegetation health and productivity over time. It also helps to map and classify land cover types, such as forests, croplands, and grasslands. NDVI values tend to be greater for areas with healthy, green vegetation and less for areas with little or no vegetation, such as bare soil or water. Analyzing NDVI values over time can effectively monitor changes in land cover and vegetation health. This method can detect decreases in NDVI values, indicating potential issues such as drought or deforestation.
The Normalized difference water index (NDWI) is a widely used water index that utilizes satellite imagery to quantify water content and detect changes in water bodies, offering valuable insights into water resource management and monitoring [56]. NDWI is calculated based on the ratio of the green band to the near-infrared band (Equation (2)), where the index values indicate the presence and condition of water bodies.
N D W I = ( G R E E N − N I R )   /   ( G R E E N + N I R )
Brightness temperature (Equation (3)) measures the apparent temperature of an object or the Earth’s surface as observed by a satellite sensor. This step followed the methods outlined on the USGS website [57].
B r i g h t n e s s   t e m p e r a t u r e ,   b t = K 2 l n ⁡ K 1 L λ + 1 − 273.15
Where, K 1 and K 2 are the thermal conversion constant and L λ is Top of Atmospheric reflectance. K 1 and K 2 can be found in the satellite metadata.
The term “proportion of vegetation” typically refers to the percentage or ratio of an area covered by vegetation in a specific land area or pixel. This metric provides insights into the presence and density of vegetation by utilizing indices or measurements. A commonly employed method for determining the proportion of vegetation in remote sensing and satellite imagery is the NDVI. Quintano et al. [58] have documented a methodology(Equations (4) and (5)) to quantify the proportion of vegetation using NDVI data.
t h e   p r o p o r t i o n   o f   v e g e t a t i o n , P v = N D V I − N D V I m i n N D V I m a x − N D V I _ m i n   2
The emissivity was calculated using the approach outlined by Sobrino et al.[59].
t h e   p r o p o r t i o n   o f   v e g e t a t i o n , E = m × P v + n  
The process of converting brightness temperature (Equation (6)) to actual land surface temperature was performed using the methodology outlined in the studies by Chander et al. [60] and Sobrino et al. [59].
L a n d   s u r f a c e   T e m p e r a t u r e , L S T = b t 1 + [ w × b t p × ln ⁡ E ]
Where:
(w): length of emitted radiation (band 10), with a value of 0.00115
(p): a constant value derived from h × c / σ = 1.438 × 10 2 , where (h): Plank’s constant (6.626 × 10-34 m2 kg/s), (c) the velocity of light ( 3 × 10 8 m / s ), ( σ ) the Boltzmann constant (1.38 × 10-23 J/K)

2.4. LULC Modeling

The original dataset originally contained 35 classes, which were subsequently consolidated into 7 distinct categories for this study. This categorization process involved reclassifying the initial classes through an image analysis method. Table 2 provides detailed descriptions of these 7 consolidated LULC classes.
Using the MOLUSC plugin within QGIS, we implemented a CA-ANN model to predict future LULC events. The Euclidean distance tool in ArcGIS v2.8 [61] was used to determine the distance from major highways. The predictive model uses two LULC raster datasets from 2 time periods as dependent and independent variables. This analysis integrated key factors such as elevation, slope, NDVI, NDWI, LST, and distance from major highways, as land-use factors. We evaluated potential transitions in LULC using conversion metrics and change probabilities obtained from the input variables. The ANN model was used to predict these changes, and subsequently, CA techniques were used to generate simulated maps. For example, classified images from 2000 and 2010 were used to predict the LULC of 2020, followed by validation against actual 2020 classified images. This method was extended to simulate LULC patterns for 2030, 2060, and 2100, establishing satisfactory accuracy. The CA model is defined by seethe mathematical equation written as Equation (7) (Figure 2 for detailed process).
S   t , t + 1 = f ( s t , N
Where, S (t+1) denotes the condition of the system at moment (t, t+1). Based on a bottom-up method, this technique analyses accurate models of urban growth and LULC processes [62].

2.5. Simulation Validation

Validation of simulated LULC is essential to affirm the accuracy and reliability of LULC models. For the year 2020, validation was conducted by comparing simulated LULC data with reference datasets using metrics such as Overall Correctness and Kappa statistics (Equations (8)–(10)) [30,63,64]. Kappa (overall) measures the agreement between observed and predicted classifications, accounting for chance agreement, with higher values indicating better predictive accuracy. Kappa Location evaluates spatial accuracy by evaluating positional agreement between observed and predicted classifications. This metric helps in understanding geographic accuracy. Kappa Histogram assesses accuracy by comparing category distributions, highlighting how well the model predicts category proportions and provides insight into quantity agreement.
K = P A − P E 1 − P E
K l o c = P A − P E P m a x − P E
K h = P m a x − P E 1 − P E
W h e r e ,   P A = ∑ i = 1 c p i i ,   P E = ∑ i = 1 c p i T p T i ,   P m a x = ∑ i = 1 c m i n p i T , p T i
a n d   ( p i j )   i s   t h e   ( i , j )   t h   c e l l   o f   t h e   c o n t i n g e n c y   t a b l e .
p i T   r e p r e s e n t s   t h e   s u m   o f   a l   l c e l l s   i n   t h e i   t h   r o w ,
p T j   r e p r e s e n t s   t h e   s u m   o f   a l l   c e l l s   i n   t h e   (   j   ) t h   c o l u m n ,
a n d   (   c   )   i s   t h e   c o u n t   o f   r a s t e r   c a t e g o r i e s

3. Result

3.1. Spatial-Temporal Distributions of Land Surface Temperature and Vegetation Coverage

Key spatial variables such as NDVI, LST, NDWI, elevation, and slope demonstrate notable variations and trends for the study area during the 2000 -2020 study period. (Table 4). In 2000, the NDVI ranged from a high of 0.52 to a low of -0.54, with an average of 0.17 and a standard deviation of 0.15 (Table 4). By 2010, the NDVI reached a maximum of 0.53 and a minimum of -0.44, with an average of 0.14 and a standard deviation of 0.14. In 2020, the NDVI varied between a high of 0.51 and a low of -0.17, averaging 0.17 with a standard deviation of 0.15. In terms of LST, the highest temperature recorded in 2000 was 31.62°C, while the lowest was 22.48°C, averaging 27.02°C. For 2010 and 2020, the maximum LSTs were 32.44°C and 32.48°C, respectively, with minimum temperatures of 23.52°C and 22.96°C, and averages of 27.95°C and 27.78°C. In 2000, the NDWI fluctuated from a high of 0.73 to a low of -0.69, with an average value of -0.19. By 2010 and 2020, the NDWI remained relatively stable, with maximums of 0.58 and 0.59, minimums of -0.64 and -0.71, and averages of -0.15 and -0.24, respectively. The elevation ranged from -41 m to 179 m, with an average of 6.46 m, and the slope varied from 0.93 to 53, averaging 2.48.
Preprints 111177 g003
Figure 3. Key LULC Variable map. Here a- f indicates the Average of NDVI, NDWI, LST, Elevation, slope, and Distance matrix variables during 2000-2020.
Figure 3. Key LULC Variable map. Here a- f indicates the Average of NDVI, NDWI, LST, Elevation, slope, and Distance matrix variables during 2000-2020.
Preprints 111177 g003
The trend of forest loss has exhibited fluctuations over time, with an overall increasing trend. From 2003 to 2012, forest loss values fluctuated periodically. Subsequently, there was a steady increase observed from 2012 to 2019, characterized by minor variations. Notably, forest loss spiked in 2020, followed by slight declines in 2021 and 2022. However, in 2023, forest loss escalated to unprecedented levels, surpassing the previous peak observed in 2020. The highest recorded values occurred in 2023, 2020, 2019, 2018, and 2016, while the lowest values were noted in 2001, 2002, and 2007 (Figure 4). This loss can be primarily attributed to unsustainable and illegal logging, unresolved land disputes, and agricultural development. The impact of these factors on forest loss will be discussed in a later section.

3.2. Spatial-Temporal Distributions of LULC

Between 1990 and 2000, significant changes occurred in LULC classes. Notably, mangrove areas decreased by 104.25 km2, from 1471.71 to 1367.46 km2 mostly in the southern part, while barren land expanded substantially by 31.09 km2; increasing from 0.32 to 31.4 km2 (Figure 5 and Table 5). Meanwhile, sparse vegetation areas experienced a significant growth of 528.22 km2, from 770.67 to 1298.89 km2, whereas dense forest areas contracted by 258.33 km2, decreasing from 932.67 to 674.34 km2. Additionally, cropland areas increased by 161.55 km², from 10,550.82 to 10,712.37 km², as cropland displaced dense forest and mangrove areas. Waterbody areas decreased by 374.42 km², from 3,819.16 km² to 3,444.74 km². Furthermore, impervious surfaces, predominantly urban areas, expanded by 16.14 km², from 39.58 to 55.72 km². Between 2000 and 2010, mangrove areas continued to decline, decreasing by 33.05 km², from 1,367.46 to 1,334.41 km² while bare surfaces expanded further by 5.06 km2, from 31.4 to 36.46 km². Sparse vegetation areas experienced significant growth of 161.5 km², increasing from 1,298.89 to 1,460.39 km² and dense forest areas increased by 41.28 km², expanding from 674.34 to 715.62 km². However, cropland decreased by 75.01 km², from 10,712.37 to 10,637.36 km2, and waterbody decreased by 122.59 km², from 3,444.74 to 3,322.15 km². Moreover, impervious surfaces expanded by 22.82 km², from 55.72 to 78.54 km². Between 2010 and 2020, mangroves continued to decline by 51.47 km², from 1,334.41 to 1,282.93 km², while bare surfaces decreased by 1.65 km², from 36.46 to 34.8 km². Sparse vegetation experienced a growth of 78.42 km², increasing from 1,460.39 to 1,538.81 km², whereas dense forest decreased by 105.02 km², from 715.62 to 610.6 km². Cropland decreased by 18.93 km2, from 10,637.35 to 10,618.42 km², and waterbody increased by 87.1 km², from 3,322.15 to 3,409.24 km². Lastly, impervious surfaces expanded by 11.56 km², rising from 78.54 to 90.1 km² (Table 5, Figure 5 and Figure 7).

3.3. Smulation of LULC

Accuracy assessment for the simulation was conducted by comparing the classified 2020 LULC with the simulated 2020 LULC. The results showed a high percentage of agreement at 90.55%, indicating strong overall model performance. The Kappa statistic (overall) was 0.83891, reflecting substantial agreement between observed and predicted classifications. Additionally, the Kappa histogram and Kappa location values were 0.97261 and 0.86253, respectively, further demonstrating the model's reliability and precision.
Predicted LULC trends reveal cropland as the predominant land cover in the Delta region, albeit with a slight decrease over time; while waterbodies consistently rank second with minor fluctuations. Mangrove coverage shows a notable decline, projected to decrease steadily from 2020 to 2100. Specifically, mangroves are projected to decrease by 13.58 km2 between 2020 and 2030, with an additional decline of 1.19 km2 from 2060 to 2100. In contrast, cropland shows a steady increase, expanding by 22.65 km2 between 2020 and 2030, and continuing to grow slightly through 2100 (Figure 6 and Figure 7).

4. Discussion

This study investigated LULC dynamics in the Irrawaddy Delta and revealed significant reductions in mangrove coverage and increases in cropland between 1990 and 2020. Additionally, the study utilized a CA-ANN model to simulate future LULC changes from 2030 to 2100, incorporating LULC variables and the RCPs future scenarios. The significant decrease in mangrove areas and dense forest areas between 1990 and 2000 is a concern, as mangroves provide crucial coastal protection, sediment and nutrient traps, and habitat for marine life [65,66]. The continued decline of mangrove forests between 2000 and 2010 (33.05 km2) highlights the need for conservation efforts [67]. In agreement with Webb et al., [5] and Yang et al., [68] data showed that agricultural expansion has been the primary cause of mangrove loss in the Delta over the past 50 years, with aquaculture and salt farming also contributing significantly. The shift to a market-oriented economy and privatization has further accelerated deforestation efforts. However, the Meinmahla Kyun Wildlife Sanctuary remains largely intact, reflecting the effectiveness of protection measures like the Association of Southeast Asian Nations (ASEAN) Heritage Park designation and the Ramsar Convention on Wetlands. The loss of mangroves in the Irrawaddy River delta is a complex issue with multiple contributing factors. Xiong [69] highlights the role of human activities, particularly deforestation and aquaculture, in the decline of mangrove forests. Jones [70] emphasizes the impact of upstream fluvial processes on mangrove sedimentation, with deforestation and urbanization increasing sedimentation rates and dams and flow diversion decreasing sediment influx. Sirisena [11] adds to this by discussing the projected future changes in streamflow and sediment loads in the Irrawaddy River Basin, with planned reservoirs potentially exacerbating the issue. Moreover, the loss of mangroves in the Irrawaddy River delta might be closely related to changes in river flux. The reduced sediment load caused by upstream developments and altered streamflow diminishes the nutrient supply necessary for mangrove growth. This not only weakens the mangrove ecosystems but also reduces their ability to protect coastal areas from erosion and storm surges, exacerbating the vulnerability of local communities to natural disasters and further disrupting the ecological balance of the delta. While Webb et al. [5] forecast complete mangrove deforestation by 2035 but our finding did not find complete deforestation of mangrove. The loss of mangroves has reduced biodiversity and their role as bio shields, increasing the risk of damage from tropical cyclones and flooding, particularly in low-lying areas of the Delta. The simulated future mangrove coverage also showed a decreasing trend (Figure 7), which poses several ecological and socio-economic challenges. Furthermore, mangroves serve as crucial habitats for numerous marine species, supporting biodiversity and local fisheries [71]. The loss of mangroves could lead to declines in fish populations and other marine life, impacting the livelihoods of local communities that depend on these resources. Additionally, mangroves play a significant role in carbon sequestration, helping to mitigate climate change [72]. The reduction in mangrove areas may therefore contribute to higher atmospheric CO2 levels, exacerbating global warming. Studies such as Chen et al.[8] and Xiong et al. [69] indicate aquaculture and rice agriculture are the main drivers of mangrove loss. Addressing these drivers involves implementing sustainable practices in aquaculture and agriculture to minimize their environmental impact. For instance, promoting integrated mangrove-aquaculture systems, where mangroves are conserved or even replanted alongside aquaculture ponds, can help maintain ecological functions while supporting local livelihoods [73,74]. Additionally, adopting agroforestry techniques in rice cultivation can reduce the need for extensive land conversion and enhance biodiversity[75]. The substantial growth of sparse vegetation areas (528.22 km2) could be a result of reforestation efforts or natural regeneration [24,76]. The contraction of dense forest areas (258.33 km2) is alarming, as these ecosystems are essential for biodiversity and carbon sequestration [77]. Previous studies such as Adas and Vogel et al. [78,79] showed that agriculture has dominated the Delta since 1974, with a significant shift from dry crops to irrigation, increasing irrigated areas from 24% to 50% by 2021. This transformation began in the 19th century under British colonialization, converting the region into a major rice-exporting area. Additionally, this study found that the increase in cropland areas by 161.55 km² may be due to agricultural expansion and food security needs [80]. The decrease in waterbody areas (3819 km2 to 3409 km2 between 1990 and 2100) is worrisome, as water resources are vital for human use, agriculture, and ecosystems [81,82]. The expansion of bare and impervious surfaces in the Delta is a result of various factors, including urbanization, infrastructure development, and natural disasters [78,83,84]. These changes have been identified and delineated through the analysis of satellite images, which have also revealed potential disaster risk drivers such as urban growth, mangrove deforestation, and the expansion of agricultural areas [78]. The impact of anthropogenic activities, including mining, on the delta's evolution has been highlighted, raising concerns about the potential disturbance of the delta's natural equilibrium [8,85]. This anticipated growth could further exacerbate existing environmental issues such as increased runoff and flood risks, the urban heat island effect, and the loss of critical ecosystems. Sustainable urban planning and environmental conservation are essential to mitigate these impacts and support the Irrawaddy Delta's ecological and socio-economic well-being. However, between 1990 and 2020, the growth of sparse and vegetation areas has minimally increased in this region, which could be attributed to natural regeneration [67]. This increase in sparse vegetation may provide some ecological benefits, such as soil stabilization and habitat for certain wildlife species, potentially offsetting some of the negative impacts of urbanization and land conversion. Nonetheless, the scattered nature of this vegetation growth suggests that it may not be sufficient to fully counterbalance the extensive loss of dense mangrove forests and other critical ecosystems.
The calculation of forest loss in the Irrawaddy Delta showed significant fluctuations and an overall increasing trend over the past two decades. From 2003 to 2012, forest loss oscillated due to natural events like Cyclone Nargis in 2008, which caused extensive mangrove damage [86], and varying levels of illegal logging and agricultural practices [87]. In this region, mangroves are mostly affected by dense human activities [86]. Between 2012 and 2019, forest loss increased gradually, driven by agricultural expansion, aquaculture development, and urbanization pressures [88]. The significant spike in 2020 can be attributed to weakened environmental regulation enforcement amid political instability and economic challenges exacerbated by the COVID-19 pandemic [89]. In 2023, forest loss surged to unprecedented levels, potentially due to intensified agricultural expansion [69], increased extreme weather events linked to climate change [90], and socioeconomic pressures on local communities to exploit forest resources for survival [91]. These dynamics highlight the complex interplay of environmental, economic, and policy factors affecting forest conservation in the region.

5. Conclusions

The Irrawaddy Delta has undergone significant LULC changes over the recent decades, driven primarily by agricultural expansion, urbanization, and climate change impacts. From 1990 to 2020, this study reveals a significant decrease in mangrove forests and an increase in cropland (rice paddies). Predictions indicate these trends will continue, with further mangrove reductions and slight fluctuations in cropland by 2100. These findings highlight the adverse effects of human activities and natural events, such as cyclones, on the region's ecological balance and socio-economic stability. Future research directions should prioritize the development of comprehensive models that incorporate socioeconomic factors, policy changes, and climate scenarios to enhance the accuracy of LULC change predictions. Additionally, there is a critical need for studies assessing the effectiveness of conservation strategies and sustainable land management practices specific to the Delta. Initiatives focusing on mangrove forest restoration and promoting sustainable agricultural practices are key. Collaborative efforts involving local communities, policymakers, and international organizations are fundamental to devising and implementing strategies aimed at enhancing the resilience of the Irrawaddy Delta's ecosystems and supporting local livelihoods. Advancements in remote sensing technologies and data collection methodologies will play a key role in providing accurate and timely information to guide these conservation and management efforts effectively.

Author Contributions

All authors contributed equally, and all authors have read and agreed to the published version of the manuscript.

Funding

This research was not funded by any specific grant from public, commercial, or not-for-profit funding sources.

Data Availability Statement

Data will be made available upon request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Syvitski, J.P.M.; Kettner, A.J.; Overeem, I.; Hutton, E.W.H.; Hannon, M.T.; Brakenridge, G.R.; Day, J.; Vörösmarty, C.; Saito, Y.; Giosan, L.; et al. Sinking Deltas Due to Human Activities. Nat. Geosci. 2009, 2, 681–686.
  2. Vörösmarty, C.J.; Syvitski, J.; Day, J.; Sherbinin, A.D.; Giosan, L.; Paola, C. Battling to Save the World’s River Deltas. Bulletin of the Atomic Scientists 2009, 65, 31–43.
  3. Renaud, F.G.; Syvitski, J.P.; Sebesvari, Z.; E Werners, S.; Kremer, H.; Kuenzer, C.; Ramesh, R.; Jeuken, A.; Friedrich, J. Tipping from the Holocene to the Anthropocene: How threatened are major world deltas?. Curr. Opin. Environ. Sustain. 2013, 5, 644–654, . [CrossRef]
  4. Shou-Hao, C. Monitoring and Predicting the Change of Mangrove Forests in Irrawaddy Delta, South Myanmar.; Citeseer: Quezon City, Metro Manila, 2015.
  5. Webb, E.L.; Jachowski, N.R.; Phelps, J.; Friess, D.A.; Than, M.M.; Ziegler, A.D. Deforestation in the Ayeyarwady Delta and the conservation implications of an internationally-engaged Myanmar. Glob. Environ. Chang. 2014, 24, 321–333, . [CrossRef]
  6. Besset, M.; Anthony, E.J.; Dussouillez, P.; Goichot, M. The Impact of Cyclone Nargis on the Ayeyarwady (Irrawaddy) River Delta Shoreline and Nearshore Zone (Myanmar): Towards Degraded Delta Resilience? Comptes Rendus Geoscience 2017, 349, 238–247.
  7. Win, S.; Towprayoon, S.; Chidthisong, A. Mangrove status, its ecosystem, and climate change in Myanmar: A study in Ayeyarwaddy Delta Coastal Zone. IOP Conf. Series: Earth Environ. Sci. 2020, 496, . [CrossRef]
  8. Chen, D.; Li, X.; Saito, Y.; Liu, J.P.; Duan, Y.; Liu, S.; Zhang, L. Recent evolution of the Irrawaddy (Ayeyarwady) Delta and the impacts of anthropogenic activities: A review and remote sensing survey. Geomorphology 2020, 365, 107231, . [CrossRef]
  9. Sakai, T.; Omori, K.; Oo, A.N.; Zaw, Y.N. Monitoring saline intrusion in the Ayeyarwady Delta, Myanmar, using data from the Sentinel-2 satellite mission. Paddy Water Environ. 2021, 19, 283–294, . [CrossRef]
  10. Salmivaara, A.; Kummu, M.; Keskinen, M.; Varis, O. Using Global Datasets to Create Environmental Profiles for Data-Poor Regions: A Case from the Irrawaddy and Salween River Basins. Environ. Manag. 2013, 51, 897–911, . [CrossRef]
  11. Sirisena, T.; Maskey, S.; Bamunawala, J.; Coppola, E.; Ranasinghe, R. Projected Streamflow and Sediment Supply under Changing Climate to the Coast of the Kalu River Basin in Tropical Sri Lanka over the 21st Century. Water 2021, 13, 3031, . [CrossRef]
  12. E Dunn, F.; E Darby, S.; Nicholls, R.J.; Cohen, S.; Zarfl, C.; Fekete, B.M. Projections of declining fluvial sediment delivery to major deltas worldwide in response to climate change and anthropogenic stress. Environ. Res. Lett. 2019, 14, 084034, . [CrossRef]
  13. Eckstein, D.; Künzel, V.; Schäfer, L. Global Climate Risk Index 2021. Who Suffers Most from Extreme Weather Events 2021, 2000–2019.
  14. Ranasinghe, R.; Wu, C.S.; Conallin, J.; Duong, T.M.; Anthony, E.J. Disentangling the relative impacts of climate change and human activities on fluvial sediment supply to the coast by the world’s large rivers: Pearl River Basin, China. Sci. Rep. 2019, 9, 1–10, . [CrossRef]
  15. GCCA Myanmar Climate Chnage Alliance 2012.
  16. Shrestha, S.; Lohpaisankrit, W. Flood hazard assessment under climate change scenarios in the Yang River Basin, Thailand. Int. J. Sustain. Built Environ. 2017, 6, 285–298, . [CrossRef]
  17. Taft, L.; Evers, M. A Review of Current and Possible Future Human-Water Interactions in Myanmar’s River Basins. Hydrol. Earth Syst. Sci. Discuss 2016, 1–37.
  18. Aung, M.T.; Shrestha, S.; Weesakul, S.; Shrestha, P.K.; others Multi-Model Climate Change Projections for Belu River Basin, Myanmar under Representative Concentration Pathways. Journal of Earth Science & Climatic Change 2016, 7.
  19. Hennig, T. Damming the Transnational Ayeyarwady Basin. Hydropower and the Water-Energy Nexus. Renewable and Sustainable Energy Reviews 2016, 65, 1232–1246.
  20. Kattelus, M.; Rahaman, M.M.; Varis, O. Hydropower development in Myanmar and its implications on regional energy cooperation. Int. J. Sustain. Soc. 2015, 7, 42, . [CrossRef]
  21. Hedley, P.J.; I Bird, M.; Robinson, R.A.J. Evolution of the Irrawaddy delta region since 1850. Geogr. J. 2010, 176, 138–149, . [CrossRef]
  22. Rai, R.; Zhang, Y.; Paudel, B.; Li, S.; Khanal, N.R. A Synthesis of Studies on Land Use and Land Cover Dynamics during 1930–2015 in Bangladesh. Sustainability 2017, 9, 1866, . [CrossRef]
  23. Renaud, F.G.; Sebesvari, Z.; Gain, A.K. Assessment of Land/Catchment Use and Degradation. In Handbook of Water Resources Management: Discourses, Concepts and Examples; Bogardi, J.J., Gupta, J., Nandalal, K.D.W., Salamé, L., Van Nooijen, R.R.P., Kumar, N., Tingsanchali, T., Bhaduri, A., Kolechkina, A.G., Eds.; Springer International Publishing: Cham, 2021; pp. 471–487 ISBN 978-3-030-60145-4.
  24. FAO Global Forest Resources Assessment-Country Report: Myanmar; 2015; p. 86;.
  25. Spalding, M.; Blasco, E.; Field, C. World Mangrove Alias. The International Society for Mangrove Ecosystems: Okinawa, Japan 1997.
  26. Maung, W. Forest Management Plan in Pyanpon District for 2006–2016. No.(1) Forest Department, Myanmar 2005.
  27. Oo, N. Present state and problems of mangrove management in Myanmar. Trees 2002, 16, 218–223, . [CrossRef]
  28. Wei, Q.; Zhiguo, P.; Jingxuan, L.; June, F.; Xiaotao, L.; Tianjie, L.; Lin, L.; Yanan, T. Remote Sensing Analysis on Ecosystem Evolution Of The Irrawaddy River Delta. In Proceedings of the IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium; IEEE: Yokohama, Japan, July 2019; pp. 6535–6538.
  29. Gilman, E.L.; Ellison, J.; Duke, N.C.; Field, C. Threats to mangroves from climate change and adaptation options: A review. Aquat. Bot. 2008, 89, 237–250, . [CrossRef]
  30. Ahmad, H.; Abdallah, M.; Jose, F.; Elzain, H.E.; Bhuyan, S.; Shoemaker, D.J.; Selvam, S. Evaluation and mapping of predicted future land use changes using hybrid models in a coastal area. Ecol. Informatics 2023, 78, 102324. [CrossRef]
  31. Ahmad, H.; Jose, F.; Shoemaker, D.J. Mapping, Dynamics, and Future Change Analysis of Sundarbans Delta Using Cellular Automata and Artificial Neural Network Modeling. IEEE J. Sel. Top. Appl. Earth Obs. Remote. Sens. 2024, 17, 5594–5603, . [CrossRef]
  32. Cabral, P.; Zamyatin, A. Markov Processes in Modeling Land Use and Land Cover Changes in Sintra-Cascais, Portugal. Dyna 2009, 76, 191–198.
  33. Kafy, A.-A.; Naim, M.N.H.; Subramanyam, G.; Ahmed, N.U.; Rakib, A.A.; Kona, M.A.; Sattar, G.S.; others Cellular Automata Approach in Dynamic Modelling of Land Cover Changes Using RapidEye Images in Dhaka, Bangladesh. Environmental Challenges 2021, 4, 100084.
  34. Saputra, M.H.; Lee, H.S. Prediction of Land Use and Land Cover Changes for North Sumatra, Indonesia, Using an Artificial-Neural-Network-Based Cellular Automaton. Sustainability 2019, 11, 3024, doi:10.3390/su11113024.
  35. Sakizadeh, M.; Milewski, A. Quantifying LULC changes in Urmia Lake Basin using machine learning techniques, intensity analysis and a combined method of cellular automata (CA) and artificial neural networks (ANN) (CA-ANN). Model. Earth Syst. Environ. 2024, 1–20, . [CrossRef]
  36. Gidey, E.; Dikinya, O.; Sebego, R.; Segosebe, E.; Zenebe, A. Cellular automata and Markov Chain (CA_Markov) model-based predictions of future land use and land cover scenarios (2015–2033) in Raya, northern Ethiopia. Model. Earth Syst. Environ. 2017, 3, 1245–1262, . [CrossRef]
  37. Asadi, M.; Oshnooei-Nooshabadi, A.; Saleh, S.-S.; Habibnezhad, F.; Sarafraz-Asbagh, S.; Van Genderen, J.L. Urban Sprawl Simulation Mapping of Urmia (Iran) by Comparison of Cellular Automata–Markov Chain and Artificial Neural Network (ANN) Modeling Approach. Sustainability 2022, 14, 15625, . [CrossRef]
  38. Xing, W.; Qian, Y.; Guan, X.; Yang, T.; Wu, H. A novel cellular automata model integrated with deep learning for dynamic spatio-temporal land use change simulation. Comput. Geosci. 2020, 137, 104430. [CrossRef]
  39. Omrani, H.; Tayyebi, A.; Pijanowski, B. Integrating the multi-label land-use concept and cellular automata with the artificial neural network-based Land Transformation Model: an integrated ML-CA-LTM modeling framework. GIScience Remote. Sens. 2017, 54, 283–304, . [CrossRef]
  40. Moskolai, W.R.; Abdou, W.; Dipanda, A.; Kolyang A Comparative Study of Deep Learning and CA-Markov Methods for Land Use / Land Cover Change Prediction. In Proceedings of the 2022 16th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS); IEEE: Dijon, France, October 2022; pp. 190–197.
  41. Furuichi, T.; Win, Z.; Wasson, R.J. Discharge and suspended sediment transport in the Ayeyarwady River, Myanmar: centennial and decadal changes. Hydrol. Process. 2009, 23, 1631–1641, . [CrossRef]
  42. Brakenridge, G.; Syvitski, J.; Niebuhr, E.; Overeem, I.; Higgins, S.; Kettner, A.; Prades, L. Design with nature: Causation and avoidance of catastrophic flooding, Myanmar. Earth-Science Rev. 2017, 165, 81–109, . [CrossRef]
  43. HIC The Ayeyarwady State of the Basin Assessmen; Yangon, 2017;
  44. Stanley, D.J.; Warne, A.G. Worldwide Initiation of Holocene Marine Deltas by Deceleration of Sea-Level Rise. Science 1994, 265, 228–231, . [CrossRef]
  45. Woodroffe, C.D. Deltaic and estuarine environments and their Late Quaternary dynamics on the Sunda and Sahul shelves. J. Asian Earth Sci. 2000, 18, 393–413, . [CrossRef]
  46. Orton, G.J.; Reading, H.G. Variability of deltaic processes in terms of sediment supply, with particular emphasis on grain size. Sedimentology 1993, 40, 475–512, . [CrossRef]
  47. Volker, A. The Deltaic Area of the Irrawaddy River in Burma. In Proceedings of the Scientific problems of the humid tropical zone deltas and their implications, Proceedings of the Dacca Symposium, UNESCO; 1966; pp. 373–379.
  48. Frenken, K. Irrigation in Southern and Eastern Asia in Figures: AQUASTAT Survey-2011. Water Reports 2012.
  49. Gordon, R. Report on the Irrawaddy River; Printed at the PW Secretariat Press, 1879;
  50. Milliman, J.D.; Meade, R.H. World-Wide Delivery of River Sediment to the Oceans. J. Geol. 1983, 91, 1–21, . [CrossRef]
  51. Velden, J. Understanding River Dynamics of the Ayeyarwady River, Myanmar How Dynamic Behaviour Contributes to Adapting the River Morphology for Navigational Purposes; 2015;
  52. Robinson, R.A.J.; Bird, M.I.; Oo, N.W.; Hoey, T.B.; Aye, M.M.; Higgitt, D.L.; X., L.X.; Swe, A.; Tun, T.; Win, S.L. The Irrawaddy River Sediment Flux to the Indian Ocean: The Original Nineteenth-Century Data Revisited. J. Geol. 2007, 115, 629–640, . [CrossRef]
  53. Hansen, M.C.; Potapov, P.V.; Moore, R.; Hancher, M.; Turubanova, S.A.; Tyukavina, A.; Thau, D.; Stehman, S.V.; Goetz, S.J.; Loveland, T.R.; et al. High-resolution global maps of 21st-century forest cover change. Science 2013, 342, 850–853, doi:10.1126/science.1244693.
  54. Zhang, X.; Liu, L.; Chen, X.; Gao, Y.; Xie, S.; Mi, J. GLC_FCS30: global land-cover product with fine classification system at 30 m using time-series Landsat imagery. Earth Syst. Sci. Data 2021, 13, 2753–2776, . [CrossRef]
  55. KRIEGLER, F.J. Preprocessing Transformations and Their Effects on Multspectral Recognition. In Proceedings of the Proceedings of the Sixth International Symposium on Remote Sesning of Environment; 1969; pp. 97–131.
  56. Gao, B.-C. NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 1996, 58, 257–266. [CrossRef]
  57. USGS Landsat 7 Data Users Handbook. 2019.
  58. Quintano, C.; Fernández-Manso, A.; Calvo, L.; Marcos, E.; Valbuena, L. Land surface temperature as potential indicator of burn severity in forest Mediterranean ecosystems. Int. J. Appl. Earth Obs. Geoinformation 2015, 36, 1–12, . [CrossRef]
  59. Sobrino, J.A.; Jiménez-Muñoz, J.C.; Paolini, L. Land surface temperature retrieval from LANDSAT TM 5. Remote. Sens. Environ. 2004, 90, 434–440, . [CrossRef]
  60. Chander, G.; Markham, B.L.; Helder, D.L. Summary of current radiometric calibration coefficients for Landsat MSS, TM, ETM+, and EO-1 ALI sensors. Remote. Sens. Environ. 2009, 113, 893–903, . [CrossRef]
  61. ESRI, I. ArcGIS Pro (Version 2.5) 2020.
  62. NextGIS MOLUSCE—Quick and Convenient Analysis of Land Cover Changes. 2017.
  63. Crowell, M.; Leatherman, S.P.; Buckley, M.K. Historical Shoreline Change: Error Analysis and Mapping Accuracy. Journal of coastal research 1991, 839–852.
  64. Wiki.gis.lab Landscape Change Analysis with MOLUSCE - Methods and Algorithms Available online: https://wiki.gis-lab.info/w/Landscape_change_analysis_with_MOLUSCE_-_methods_and_algorithms#Validation (accessed on 9 June 2024).
  65. Djalante, R. Key Assessments from the IPCC Special Report on Global Warming of 1.5 C and the Implications for the Sendai Framework for Disaster Risk Reduction. Progress in Disaster Science 2019, 1, 100001.
  66. Masson-Delmotte, V.; Zhai, P.; Pörtner, H.-O.; Roberts, D.; Skea, J.; Shukla, P.R.; Pirani, A.; Moufouma-Okia, W.; Péan, C.; Pidcock, R. Global Warming of 1.5 C. An IPCC Special Report on the impacts of global warming of 2019, 1, 93–174.
  67. Outlook, G.B. Global Biodiversity Outlook 3. In Proceedings of the Montréal, Canada: Secretariat of the Convention on Biological Diversity.(http://gbo3. cbd. int/) Phil. Trans. R. Soc. B; 2010; Vol. 9.
  68. Yang, J.; Ren, W.; Ouyang, Y.; Feng, G.; Tao, B.; Granger, J.J.; Poudel, K.P. Projection of 21st century irrigation water requirement across the Lower Mississippi Alluvial Valley. Agric. Water Manag. 2019, 217, 60–72, . [CrossRef]
  69. Xiong, Y.; Dai, Z.; Long, C.; Liang, X.; Lou, Y.; Mei, X.; Nguyen, B.A.; Cheng, J. Machine Learning-Based examination of recent mangrove forest changes in the western Irrawaddy River Delta, Southeast Asia. CATENA 2024, 234, 107601. [CrossRef]
  70. Jones, S. From Source to Sink: How Linking Upstream Fluvial Processes to Mangrove Sedimentation Can Improve Mangrove Management Strategies. Oceanogr. Fish. Open Access J. 2020, 12, 001–004, . [CrossRef]
  71. Nagelkerken, I.; Blaber, S.J.M.; Bouillon, S.; Green, P.; Haywood, M.; Kirton, L.G.; Meynecke, J.-O.; Pawlik, J.; Penrose, H.M.; Sasekumar, A.; et al. The habitat function of mangroves for terrestrial and marine fauna: A review. Aquat. Bot. 2008, 89, 155–185, doi:10.1016/j.aquabot.2007.12.007.
  72. Sandilyan, S.; Kathiresan, K. Mangrove conservation: a global perspective. Biodivers. Conserv. 2012, 21, 3523–3542, . [CrossRef]
  73. Widigdo, B.; Ilman, M.; Santoso, N.; Sulistiono; Kusmana, C.; Krisanti, M. Performance of Integrated Mangrove-Aquaculture (IMA) in District of Sidoarjo, East Java Indonesia. BIO Web Conf. 2024, 92, 01022, . [CrossRef]
  74. Vo, T.; Le, H.; Phan, D.; Tran, A.; Nguyen, Y.; Nguyen, H.; Pham, T. Towards sustainable mangrove-shrimp aquaculture through capacity building and partnership in the Mekong River Delta. APN Sci. Bull. 2024, 14, 28–37, . [CrossRef]
  75. Rahman, S.A.; Jacobsen, J.B.; Healey, J.R.; Roshetko, J.M.; Sunderland, T. Finding alternatives to swidden agriculture: does agroforestry improve livelihood options and reduce pressure on existing forest?. Agrofor. Syst. 2016, 91, 185–199, . [CrossRef]
  76. MacDicken, K.G. Global Forest Resources Assessment 2015: What, Why and How? Forest Ecology and Management 2015, 352, 3–8.
  77. Brondizio, E.; Díaz, S.M.; Settele, J.; Ngo, H.; Gueze, M.; Aumeeruddy-Thomas, Y.; Bai, X.; Geschke, A.; Molnár, Z.; Niamir, A. Assessing a Planet in Transformation: Rationale and Approach of the IPBES Global Assessment on Biodiversity and Ecosystem Services. 2019.
  78. Vogel, A.; Seeger, K.; Brill, D.; Brückner, H.; Soe, K.K.; Oo, N.W.; Aung, N.; Myint, Z.N.; Kraas, F. Identifying Land-Use Related Potential Disaster Risk Drivers in the Ayeyarwady Delta (Myanmar) during the Last 50 Years (1974–2021) Using a Hybrid Ensemble Learning Model. Remote. Sens. 2022, 14, 3568, . [CrossRef]
  79. Adas, M. The Burma Delta: Economic Development and Social Change on an Asian Rice Frontier, 1852–1941; New Perspectives in Southeast Asian Studies; University of Wisconsin Press: Madison, WI, USA, 2011;
  80. Sunderland, T.; Powell, B.; Ickowitz, A.; Foli, S.; Pinedo-Vasquez, M.; Nasi, R.; Padoch, C. Food Security and Nutrition. Center for International Forestry Research (CIFOR), Bogor, Indonesia 2013.
  81. Boretti, A.; Rosa, L. Reassessing the projections of the World Water Development Report. npj Clean Water 2019, 2, 15, doi:10.1038/s41545-019-0039-9.
  82. Shah, E.; Liebrand, J.; Vos, J.; Veldwisch, G.J.; Boelens, R. The UN World Water Development Report 2016, Water and Jobs: A Critical Review. Dev. Chang. 2018, 49, 678–691, . [CrossRef]
  83. Khaing, S.S.; Aung, S.W.Y.; Aung, S.T. Analysis of Environmental Change Detection Using Satellite Images (Case Study: Irrawaddy Delta, Myanmar). In Big Data Analysis and Deep Learning Applications; Zin, T.T., Lin, J.C.-W., Eds.; Advances in Intelligent Systems and Computing; Springer Singapore: Singapore, 2019; Vol. 744, pp. 355–363 ISBN 9789811308680.
  84. Habitat, U. Urbanization and Development: Emerging Futures. World cities report 2016, 3, 4–51.
  85. LaJeunesse Connette, K.J.; Connette, G.; Bernd, A.; Phyo, P.; Aung, K.H.; Tun, Y.L.; Thein, Z.M.; Horning, N.; Leimgruber, P.; Songer, M. Assessment of Mining Extent and Expansion in Myanmar Based on Freely-Available Satellite Imagery. Remote Sens. 2016, 8, 912, . [CrossRef]
  86. Aung, T.T.; Mochida, Y.; Than, M.M. Prediction of recovery pathways of cyclone-disturbed mangroves in the mega delta of Myanmar. For. Ecol. Manag. 2013, 293, 103–113, . [CrossRef]
  87. Lim, C.L.; Prescott, G.W.; De Alban, J.D.T.; Ziegler, A.D.; Webb, E.L. Untangling the proximate causes and underlying drivers of deforestation and forest degradation in Myanmar. Conserv. Biol. 2017, 31, 1362–1372, . [CrossRef]
  88. Yang, R.; Luo, Y.; Yang, K.; Hong, L.; Zhou, X. Analysis of Forest Deforestation and its Driving Factors in Myanmar from 1988 to 2017. Sustainability 2019, 11, 3047, . [CrossRef]
  89. Wang, Q.; Rossignoli, C.M.; Dompreh, E.B.; Su, J.; Griffiths, D.; Htoo, K.K.; Nway, H.M.; Akester, M.; Gasparatos, A. Diversification strategies have a stabilizing effect for income and food availability during livelihood shocks: Evidence from small-scale aquaculture-agriculture systems in Myanmar during the COVID-19 pandemic. Agric. Syst. 2024, 217, 103935. [CrossRef]
  90. Srinivasan, G.; Agarwal, A.; Bandara, U. Climate Change Impacts on Water Resources and Agriculture in Southeast Asia with a Focus on Thailand, Myanmar, and Cambodia. In The Role of Tropics in Climate Change; Elsevier, 2024; pp. 17–32 ISBN 978-0-323-99519-1.
  91. World Bank Myanmar Country Environmental Analysis Available online: https://www.worldbank.org/en/country/myanmar/publication/myanmar-country-environmental-analysis (accessed on 6 June 2024).
Preprints 111177 g002
Figure 2. Methods for image collection processing, classification, accuracy assessment, and simulation for the future. The green box represents the CA-ANN modeling and simulation scheme.
Figure 2. Methods for image collection processing, classification, accuracy assessment, and simulation for the future. The green box represents the CA-ANN modeling and simulation scheme.
Preprints 111177 g002
Preprints 111177 g004
Figure 4. Forest loss over Irrawaddy Delta from 2001 to 2023.
Figure 4. Forest loss over Irrawaddy Delta from 2001 to 2023.
Preprints 111177 g004
Preprints 111177 g005
Figure 5. LULC Map of Irrawaddy Delta for 1990, 2000, 2010, and 2020.
Figure 5. LULC Map of Irrawaddy Delta for 1990, 2000, 2010, and 2020.
Preprints 111177 g005
Preprints 111177 g006
Figure 6. Predicted Land Use and Land Cover maps for the years 2030 (a), 2060 (b), and 2100 (c).
Figure 6. Predicted Land Use and Land Cover maps for the years 2030 (a), 2060 (b), and 2100 (c).
Preprints 111177 g006
Preprints 111177 g007
Figure 7. Percentage coverage of all LULC classes from 1990 to 2100, with observed values represented by a solid line and predicted values represented by a dashed line. A black vertical line indicates the transition between observed and predicted data.
Figure 7. Percentage coverage of all LULC classes from 1990 to 2100, with observed values represented by a solid line and predicted values represented by a dashed line. A black vertical line indicates the transition between observed and predicted data.
Preprints 111177 g007
Table 1. Remote sensing used variables, their source, description, and availability.
Table 1. Remote sensing used variables, their source, description, and availability.
Variable Image collection / Source Tier Description Availability
LST and NDVI LANDSAT/LT05/C01/T1_TOA Landsat 5,7,8 Collection 1 Tier 1 TOA Reflectance 1992-01- 2022-05
LULC sat-io/open-datasets/GLC-FCS30D/annual Land cover dynamics at a 30-meter 1985 to 2022
Roads OpenStreetMap - -
Slope and elevation NASA / USGS / JPL-Caltech NASA SRTM Digital Elevation 30 m 2000
Forest loss Hansen/UMD/Google/USGS/NASA Hansen Global Forest Change dataset (version v1_11) 2000 to 2023
Table 2. LULC classes and their descriptions.
Table 2. LULC classes and their descriptions.
NO. Variables Description
1 Mangrove Mangrove trees
2 Bare surfaces Bare soil, beaches, tidal flat
3 Sparse vegetation Grassland, sparse vegetation, shrubland, saltmarsh
4 Dense forest Without mangrove trees
5 Cropland Crop fields
6 Waterbody Rivers, channels, creeks, ponds, canals, lakes, and ocean water, swamp, flooded flat.
7 Impervious surfaces Built up
Table 3. Variables and their descriptions.
Table 3. Variables and their descriptions.
NO. Variables Description
1 LST Land Surface Temperature
2 NDVI Normalized Difference Vegetation Index
3 NDWI Normalized Difference Water Index
4 Distance to road Representation of the distance of each cell in the raster from the nearest road.
5 Elevation Representation of the topography of a given area.
6 Slope The steepness of the land surface.
7 Future LST RCP scenario between 2021 to 2040, 2041 to 60, and 2081 to 2100
Table 4. Descriptive Statistics of Key Spatial Variables. This table presents the descriptive statistics for key spatial variables, including the Normalized Difference Vegetation Index (NDVI), Land Surface Temperature (LST), and Normalized Difference Water Index (NDWI). The statistics provided include the maximum, minimum, mean, and standard deviation (SD) for each variable.
Table 4. Descriptive Statistics of Key Spatial Variables. This table presents the descriptive statistics for key spatial variables, including the Normalized Difference Vegetation Index (NDVI), Land Surface Temperature (LST), and Normalized Difference Water Index (NDWI). The statistics provided include the maximum, minimum, mean, and standard deviation (SD) for each variable.
Variables Max Min Mean SD
NDVI 0.52 -0.38 0.16 0.15
NDWI 0.63 -0.68 -0.19 0.31
LST 32.18 22.99 27.58 1.52
Elevation 17.00 9.00 6.00 0.46
Slope 53.00 0.93 2.48 1.85
Distance 6328.36 0.00 667.22 759.54
Table 5. LULC class-wise extracted data, year-wise coverage (km²), change (km²), and percentage.
Table 5. LULC class-wise extracted data, year-wise coverage (km²), change (km²), and percentage.
Class 1990 2000 Change 1990 % 2000 %
Mangrove 1471.71 1367.46 -104.25 8.3691 7.7763
Bare surfaces 0.32 31.4 31.09 0.0018 0.1786
Sparse vegetation 770.67 1298.89 528.22 4.3826 7.3864
Dense forest 932.67 674.34 -258.33 5.3038 3.8348
Cropland 10550.82 10712.37 161.55 59.9992 60.9179
Waterbody 3819.16 3444.74 -374.42 21.7184 19.5892
Impervious surfaces 39.58 55.72 16.14 0.2251 0.3169
Class 2000 2010 Change 2000 2010%
Mangrove 1367.46 1334.41 -33.05 7.7763 7.5884
Bare surfaces 31.4 36.46 5.06 0.1786 0.2073
Sparse vegetation 1298.89 1460.39 161.5 7.3864 8.3048
Dense forest 674.34 715.62 41.28 3.8348 4.0695
Cropland 10712.37 10637.36 -75.01 60.9179 60.4913
Waterbody 3444.74 3322.15 -122.59 19.5892 18.8920
Impervious surfaces 55.72 78.54 22.82 0.3169 0.4466
Class 2010 2020 Change 2010% 2020 %
Mangrove 1334.41 1282.93 -51.47 7.5884 7.2956
Bare surfaces 36.46 34.8 -1.65 0.2073 0.1979
Sparse vegetation 1460.39 1538.81 78.42 8.3048 8.7508
Dense forest 715.62 610.6 -105.02 4.0695 3.4723
Cropland 10637.35 10618.42 -18.93 60.4913 60.3837
Waterbody 3322.15 3409.24 87.1 18.8920 19.3873
Impervious surfaces 78.54 90.1 11.56 0.4466 0.5124
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

facebook logotwitter logolinkedin logo
MDPI Initiatives
Important Links
Subscribe

Disclaimer

Terms of Use

Privacy Policy

© 2024 MDPI (Basel, Switzerland) unless otherwise stated