Next Article in Journal
An Examination of ICESat-2 Repeat Tracks for Quantifying Hurricane-Driven Changes in Forest Structure
Previous Article in Journal
Individual Urban Tree Detection from Multispectral Satellite Imagery via Point-Supervised Deep Learning
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 18
Issue 12
10.3390/rs18122022
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Unequal Burdens: Land Tenure and Agricultural Losses in the 2019 Lower Mississippi River Floods

by
Jephthah Nimoh Marfo
and
Shrinidhi Ambinakudige
*
Department of Geosciences, Mississippi State University, Mississippi State, MS 39762, USA
*
Author to whom correspondence should be addressed.
Remote Sens. 2026, 18(12), 2022; https://doi.org/10.3390/rs18122022
Submission received: 15 February 2026 / Revised: 4 May 2026 / Accepted: 12 June 2026 / Published: 17 June 2026
Browse Figures
Versions Notes

Highlights

What are the main findings?
  • Small land parcels are more vulnerable to floods in the Delta, experiencing greater proportional impacts than larger farms.
  • Soybeans were most affected by flooding, especially in May followed by corn and cotton.
What are the implications of the main findings?
  • Flood mitigation efforts should prioritize small land parcels in the Delta to reduce production losses.
  • Target adaptation strategies to flooding vulnerabilities for soybeans, corn, and cotton.

Abstract

The 2019 Mississippi River floods were among the most severe in recent U.S. history, impacting 11 states and driven by multiple tributary flood events rather than a single episode. This study focuses on the Lower Mississippi River Basin in Mississippi, examining how flood frequency interacts with land ownership patterns to influence agricultural losses in the Yazoo–Mississippi Delta. Using Sentinel-2 imagery within Google Earth Engine, land use and land cover were classified with a random forest algorithm, followed by change detection and a flood recurrence–persistence modeling framework to map and characterize inundation. Results indicate that mid-year floods (April–July) caused the greatest crop losses, particularly in soybeans (4475 ha), cotton (501 ha), and corn (546 ha). Most impacts were associated with short-duration, low-recurrence floods, which affected many structures (1812) and extensive agricultural areas due to their broad spatial reach. Small agricultural parcels (≤48 ha) experienced the highest proportional exposure across flood zones, while medium and large parcels showed comparatively lower vulnerability. These findings highlight the importance of targeted resilience and mitigation strategies that account for flood frequency, land use, and land ownership patterns across the Delta.

1. Introduction

Agricultural systems are particularly vulnerable to flood hazards due to their susceptibility to seasonal cycles and their exposure to low-lying floodplains. Flooding can result in the complete loss of crops [1], long-term soil degradation [2,3], and decreased productivity [4,5] throughout the growing season. In addition to damaging crops, floods can also affect land ownership stability. Large-scale and smallholder farmers face different levels of exposure to flooding and have varying capacities for recovery [6]. To develop effective risk management strategies and safeguard livelihoods, it is essential to understand how flooding affects agricultural land use over time and across different areas.
The Mississippi Watershed is the largest in North America. It drains 41% of the contiguous United States and two Canadian provinces. It is the third-longest river in the world, the sixth-largest in sediment yield, and the eighth-largest in discharge volume [7,8,9]. Over the past 50 years, the Mississippi River has served as a major trade and commerce hub for the United States, a significant waste disposal site, a hydroelectric power source, and a crucial source of cooling and drinking water [8,9,10,11]. Nonetheless, the Mississippi River has a past of devastating floods, with the significant flood in 1927 leading to the establishment of a comprehensive flood management system. This system is made up of levees, upstream dams and reservoirs, as well as spillways such as the Bonnet Carré Spillway located near New Orleans, which channels excess water into Lake Pontchartrain and subsequently into the Gulf of Mexico [12,13,14]. Flooding in the basin results from both natural hydrologic processes and human modification of the river system.
The 2019 Mississippi River Basin flood was one of the most severe and destructive hydrological events in recent U.S. history. It was found to be the longest-lasting flood since the 1927 Great Flood [15]. The flooding along the Mississippi River system over the years is caused by a combination of natural and human-driven factors, with human modifications to the river system being the most significant contributors [13]. Mississippi floods in 2008, 2011, and 2015–2019 highlighted storm sequences, heavy rainfall, saturated soil, and snowmelts from the Northern Great Plains influenced floods in the lower part of Mississippi [14,16,17]. Munoz et al. [13] and Pinter et al. [18] revealed that while climate change and land use changes do play roles, the substantial increase in flood levels observed in Mississippi over the past 100 to 150 years is primarily due to river engineering. Specifically, wing dikes and levees have greatly altered the way rivers manage high-flow events.
According to Munoz et al. [13], flooding from the Mississippi River has led to significant environmental degradation, including the loss of essential wetlands and changes in wildlife habitats. These floods create social hardships by displacing communities [19] and increasing health risks due to insufficient access to clean water [20]. Armstrong et al. [21] reported that the floodwaters released from the Bonnet Carré Spillway in 2019 led to a toxic algal bloom. This bloom caused the closure of 25 Gulf Coast beaches, resulted in a dolphin mortality event, and contributed to the complete collapse of shellfisheries in several areas of the Western Mississippi Sound. Economically, the impact is severe, resulting in extensive agricultural damage, increased infrastructure repair costs, and heightened navigation expenses, all of which undermine regional trade and economic stability. For instance, in 2011, the flooding of the lower Mississippi River resulted in approximately $3.2 billion in damages to agriculture and infrastructure [12]. The 2019 Mississippi River Basin floods caused major disruptions to agricultural production [22], resulting in farmers being unable to plant on millions of acres. This led to significant financial losses, amounting to billions of dollars, and a considerable amount of unshipped grain, estimated at billions of dollars [23]. As of August 2019, John [24] reported that an estimate of about 645,800 acres of land was prevented from being planted during the flood events in Mississippi State alone. Damages in the Mississippi and Missouri River valleys surpassed $1 billion across agriculture, manufacturing, and transportation [17].
The characterization of flood frequency and extent at high temporal and spatial resolutions has been enabled by advances in remote sensing and geospatial analytics [25,26]. These capabilities have been built upon in previous studies, which have used climate models [27,28,29], satellite-derived remote sensing data [30,31,32,33], machine learning techniques [34,35,36], hydrological models [37,38,39,40,41,42], and a combination of these diverse sources and techniques to model and map flood occurrences, frequency, susceptibility, and risk. Accurate flood mapping in the Mississippi Delta requires remote sensing approaches capable of capturing spatial and temporal inundation dynamics. Synthetic Aperture Radar (SAR) is widely used for flood monitoring because it can acquire data under all weather and illumination conditions [43]. Operating both day and night, SAR is particularly valuable for detecting flooding in urban areas [44]. Flooded surfaces typically exhibit low backscatter relative to surrounding terrain, enabling rapid delineation of inundation. However, accurately mapping floods remains challenging, especially at local scales. SAR-based flood delineation may lack spatial detail due to limitations in resolution and viewing geometry, particularly in areas with complex topography, dense vegetation, or built environments [43]. In urban settings, interpretation of SAR signals is further complicated by building shadowing, multiple reflections (double-bounce), and signal interactions with flat roofs and windows, which can reduce mapping accuracy [45]. To enhance reliability and accuracy, SAR data are often integrated with other datasets and remote sensing approaches [43].
Despite the advantages of SAR, optical satellite imagery datasets remain valuable for mapping the extent of floods and monitoring changes in large flood plains, particularly when multi-temporal imagery is used to observe inundation patterns and their impact on land use [46]. The integration of spatial data is recognized as an effective means of assessing flood risk distribution and aids the high-resolution reanalysis of past flood events [47,48]. Sentinel-2 multispectral imagery is especially suitable for flood mapping due to its high spatial resolution (10–20 m), frequent revisit time, and spectral capability to distinguish water, vegetation, and soil moisture. However, spectral indices alone may misclassify shadows, wet soils, and built-up areas. Therefore, following recent studies [49,50], this study adopts a multi-layer approach by integrating Sentinel-2 mosaics with spectral indices, topographic variables, and soil-related layers to improve inundation detection. This integration improves flood detection by combining water-sensitive spectral information with terrain and soil characteristics, thereby reducing classification errors and providing a reliable method for large-scale flood mapping.
Land ownership plays a critical role in shaping exposure, vulnerability, and recovery from floods [51]. Patterns of land ownership influence how agricultural land is managed, the types of crops cultivated, and the financial capacity to withstand or adapt to repeated flood events [52]. In the Mississippi River Basin, parcel size and tenure vary widely, ranging from small family-owned farms to large, corporate-owned commercial operations [53]. These differences contribute to uneven distributions of risk, with some landowners experiencing recurrent losses while others remain relatively protected. Despite their importance, parcel-level ownership and tenure are often underutilized in flood impact studies, limiting our understanding of how floods reshape agricultural systems at the scale where decisions are made and adaptations occur.
Despite substantial progress in flood mapping through remote sensing, hydrological modeling, and geospatial analytics, several important research gaps remain. Previous studies [27,54,55] have primarily focused on delineating flood extent, forecasting inundation, or assessing hazard susceptibility at regional scales, with limited attention to the temporal behavior of floods, such as persistence, recurrence, and duration at fine spatial resolutions. In agricultural floodplains, these temporal dynamics are critical because short-term widespread flooding and prolonged localized inundation can produce very different impacts on crops, infrastructure, and land productivity. In addition, relatively few studies have linked flood exposure to parcel-level land ownership and tenure patterns, even though ownership structure strongly influences management capacity, recovery potential, and access to mitigation resources. Crop-specific vulnerability is also insufficiently addressed, particularly in regions where multiple crops occupy landscapes with varying flood sensitivities. These gaps are especially important in the Mississippi Delta, where repeated flooding intersects with intensive agriculture and diverse landholding patterns. Addressing these limitations requires an integrated framework that combines flood dynamics, agricultural exposure, and land ownership characteristics to better support equitable and effective flood-risk management.
To address these gaps, this study develops an integrated geospatial framework to evaluate the 2019 Mississippi River Basin flood impacts in the Mississippi Delta using Sentinel-2 imagery, spectral indices, topographic variables, soil-related datasets, cadastral parcel information, and crop distribution data. The specific objectives of the study are to: (1) map flood extent, persistence, and recurrence at high spatial resolution for the January–August 2019 flood season; (2) quantify parcel-level flood exposure and examine how flood impacts vary among small, medium, and large agricultural land parcels; (3) estimate crop-specific agricultural losses and evaluate the temporal distribution of flood impacts across major crops, including soybeans, corn, cotton, rice, and wheat; and (4) identify spatial inequalities in socio-agricultural vulnerability by linking flood dynamics with land tenure and cultivated land intensity. By integrating flood dynamics, agricultural exposure, and ownership structure, this study provides a practical decision-support framework for flood mitigation, crop insurance prioritization, land-use planning, and resilience strategies in flood-prone agricultural landscapes.

2. Materials and Methods

2.1. Study Area

The study included 7 counties (Figure 1): Washington County, Holmes County, Humphreys County, Sharkey County, Yazoo County, Issaquena County, and Warren County. These areas are located in the Yazoo–Mississippi Delta, often referred to as the Lower Delta, and form a distinct Alluvial plain in northwestern Mississippi bounded by the Mississippi and Yazoo Rivers [56]. The Delta was historically a swampy area. It has been transformed through clearing, draining, and constructing levees for agricultural use [57]. Characterized by flat topography [58], fertile soils [59], and extensive networks of rivers and tributaries [60], the Yazoo–Mississippi Delta area is among the most agriculturally productive landscapes in the United States [61]. The Delta’s subtropical climate, marked by warm summers and frequent rainfall, supports intensive cultivation of crops such as cotton, soybeans, rice, and corn [62,63].

2.2. Data Sources

This study used a combination of remotely sensed imagery and additional geospatial data to map flood frequency, evaluate agricultural losses, and analyze land tenure patterns in the study area as described in Table 1.

2.3. Land Use Land Cover Mapping

The methodological framework employed in this study is summarized in the flowchart (Figure 2), which outlines the sequential steps for flood recurrence, persistence, and flood zone analysis, as well as infrastructure, crop land, and land parcel exposure assessment.
Sentinel-2 surface reflectance images (COPERNICUS/S2_SR) were accessed through Google Earth Engine (GEE) for the study period, filtered for <15% cloud cover within the study area. After atmospheric correction, cloud masking, and all spectral bands were scaled to surface reflectance values. Spectral indices were computed, including the Normalized Difference Vegetation Index (NDVI) [64], Normalized Difference Water Index (NDWI) [65], Normalized Difference Moisture Index (NDMI) [66], Normalized Difference Built-up Index (NDBI) [67], Automated Water Extraction Index (AWEI) [68], Bare Soil Index (BSI) [69], and Water Ratio Index (WRI) [70]. These indices were included to improve class separability and enhance discrimination among land cover types. In addition, the Optical Trapezoid Model (OPTRAM) soil moisture index [71,72] was calculated using NDVI and shortwave infrared (SWIR) reflectance, following the methodology of [72], with calibrated parameters. The OPTRAM-derived soil moisture layer was incorporated as an additional predictor variable, particularly to improve discrimination between the land cover classes. Topographic variables were incorporated to enhance classification performance. Elevation data were obtained from the USGS 3DEP Digital Elevation Model at 10 m spatial resolution. Slope layer was derived from the elevation dataset using terrain analysis functions (built-in ee.Terrain functions). Both elevation and slope were included as predictor variables in the classification process to account for terrain-driven variations in land cover distribution. A composite image was generated to include ten spectral bands (Blue, Green, Red, Red-edge bands, Near-Infrared, and Shortwave Infrared bands), slope, OPTRAM, spectral indices, and elevation at 10 m spatial resolution.
The 2018 and 2019 Cropland Data Layers (CDLs) from USDA-NASS were obtained at 30 m resolution and clipped to the study area. CDL crop codes were reclassified into six thematic classes (Table 2): Water Bodies, Agriculture, Forest, Shrubs/Grassland, Built-up, and Barren. To improve the reliability of training data and reduce noise, a spatial homogeneity filtering approach was applied. A 3 × 3 majority neighborhood-based mode filter was applied, followed by a connected pixel filter to retain only homogeneous regions with ≥50 connected pixels in the contiguous clusters. This generated “pure pixels” with at least 75% class purity, which were used as training samples following a similar methodology from Hao et al. [73] and Tran et al. [74]. This approach ensured the selection of spectrally pure training samples and minimized mixed-pixel effects, thereby improving accuracy. A stratified random sampling approach was employed to generate 2400 training samples (400 per class) to balance the representation across all six land cover categories. The spectral bands, spectral indices, OPTRAM soil moisture, elevation, and slope values were assigned to each sample region. The dataset was randomly split into training (70%) and testing (30%) subsets. The training subset was used to develop the classification model, while the validation subset was held out for independent accuracy assessment of the final classification.
Land cover classification was performed using the Random Forest algorithm implemented in GEE. To perform hyperparameter tuning, the 70% training subset was further partitioned into two groups: 80% for parameter tuning (model fitting) and 20% for internal validation. A grid search hyperparameter tuning was carried out using the 80% subset training sample by testing multiple combinations of the number of trees (100–500) and the number of variables per split (mtry: 1–7). The optimal configuration was identified as 400 trees and mtry = 7, based on accuracy and Cohen’s Kappa using the 20% subset training sample. The final classifier was applied to the final composite image to generate the final land cover map at 10 m spatial resolution.
The classification was validated using the 30% test dataset. A confusion matrix was generated to compute Overall Accuracy (OA), Producer’s Accuracy (PA), User’s Accuracy (UA), and Cohen’s Kappa coefficient (κ), as explained in many studies [75,76,77,78,79].

2.4. Change Detection and Flood Mapping

ArcGIS Pro 3.3 was used to perform a post-classification change detection between the pre-flood reference period (June 2018) and each flood month (January, March, April, May, July, and August 2019) to identify land–water transitions. A baseline permanent water mask was first generated using the USDA-NASS dataset and the June 2018 classification, representing persistent water bodies such as the Mississippi River, lakes, reservoirs, ponds, and streams. These permanent water features were excluded from subsequent analyses to prevent confusion with seasonal or inundation-related water dynamics. Flooded (inundated) areas were defined as pixels classified as non-water in the June 2018 reference image that transitioned to water in the corresponding 2019 flood-month image. Binary flood maps were generated for each time step (e.g., June 2018–January 2019, June 2018–March 2019, etc.), where inundation was identified as a change from non-water to water after masking permanent water bodies. Unchanged non-water areas and permanent water bodies were excluded from flood extent estimates to ensure that only newly inundated areas were mapped.
The accuracy of the change detection maps was evaluated using a post-classification comparison approach. Since change detection was derived from classified land cover maps, the reliability of change results depends on the accuracy of the individual classifications. Therefore, the classified maps used for change detection were first independently validated using a confusion matrix-based accuracy assessment as described in Section 2.3.
The monthly binary maps were combined using a raster calculation that summed the number of times each pixel was inundated over the study period. The resulting map showed flood frequency; higher values indicated pixels that were repeatedly inundated. To assess flooding persistence, a water inundation frequency (WIF) [80] approach was adopted, in which the inundation frequency for each pixel was normalized to the total duration of the study period (6 months). WIF was computed as
W I F = N f N t × 100
where Nf represents the number of months in which a pixel was classified as flooded, and Nt is the total number of months in the observation period.
Based on the computed WIF values, the flooding persistence map was reclassified into four categories: non-flooded, low persistence, medium persistence, and high persistence. Pixels with no flooding (0% WIF) were categorized as 0 (non-flooded area). Pixels with WIF values ranging from 16.7% to 33.3% (1–2 months of inundation) were classified as having low persistence. WIF values defined medium persistence between 50.0% and 66.7% (3–4 months of inundation). High persistence included pixels with WIF values exceeding 83.3% (5–6 months of inundation).
Flood recurrence was evaluated by generating binary flood maps for each pair of consecutive months (January–March, March–April, April–May, May–July, and July–August). Transitions from non-flooded to flooded states were recorded as recurrence events. The binary transition maps were summed to calculate the total number of flood occurrences per pixel across all monthly intervals. The resulting map reflected the frequency of flood recurrence, indicating how often new flooding occurred over previously dry areas. Finally, the cumulative recurrence raster was reclassified into three levels: low, medium, and high. These levels represent areas with one, two, and three or more instances of flood recurrence, respectively.
Finally, a composite Flood Zone Map was generated by integrating both persistence and recurrence classes to represent combined flood dynamics. Pixels identified as non-flooded in either map were assigned to the “no flood” category. All other pixels were classified based on the combined influence of their persistence and recurrence levels. This integration differentiated flood zones into multiple classes that reflect varying degrees of flood hazard severity. Areas with high persistence and recurrence were interpreted as zones of extreme flood risk, and areas with low persistence and recurrence were interpreted as regions of occasional or short-term inundation.
To further refine the analysis, additional datasets were integrated, including built-up area polygons from Mississippi Automated Resource Information system (MARIS), land parcel data containing proprietary information (owner name, cultivated and uncultivated area, total parcel size), and the 2019 CDL from NASS to identify crops adversely affected by flooding (Rice, Corn, Winter Wheat, Soybean, and Cotton). The impact of flooding on land parcels was quantified using spatial analysis tools, including zonal statistics, frequency, summary statistics, and tabulate area in ArcGIS Pro (version 3.3). Land parcels intersecting flooded zones were classified based on their areal extent (hectares) to examine how flooding impacts vary across different parcel sizes. Parcel size classification was performed using the Jenks Natural Breaks (optimization) method, which determines class intervals by minimizing variance within classes and maximizing variance between classes. This data-driven approach identifies natural groupings inherent in the parcel size distribution. Based on the resulting thresholds, parcels were categorized into three classes: small (≤48 ha), moderate (48–112 ha), and large (>112 ha). These classes represent statistically derived groupings of parcel sizes and are independent of flood exposure. Following classification, spatial analysis was conducted for each tenure category to determine the distribution of flooded cultivated and flooded areas across the flood zones.

3. Results

3.1. LULC Classification and Assessment

Figure 3, Figure 4, Figure 5 and Figure 6 collectively illustrate the spatial and temporal dynamics of land use, flooding, and change across the study area. Figure 3 presents the land use/land cover (LULC) maps for the study period, showing the distribution of key classes: water bodies, agricultural areas, forest areas, shrubland and grassland, built-up areas, and barren areas. Figure 4 shows binary change detection maps for flood periods in 2019, highlighting areas affected by flooding, permanent water bodies, and unchanged LULC classes. Figure 5 visualizes the monthly flood extent, capturing the spatial spread of inundation during each flooding period. Finally, Figure 6 provides deeper insights into flood dynamics through three thematic maps: (a) flood persistence, indicating areas consistently flooded over time; (b) flood recurrence, showing locations repeatedly affected by flooding; and (c) flood zone classification, delineating zones based on flood frequency and duration. The accuracy assessment for the LULC classification across the study period shows consistently high performance. Overall Accuracy (OA) ranged from 0.88 to 0.94 (Table S15), while the Kappa coefficient (κ), which measures agreement beyond chance, ranged from 0.85 to 0.92 (Table S15). In addition to these overall metrics, class-specific Producer’s Accuracy (PA) and Consumer’s Accuracy (CA) were computed for each classification period (Tables S1–S14). These metrics suggest that the LULC classifications were robust and reliable throughout the study period.

3.2. Flood Impacts on Built Infrastructure

July was the month with the highest frequency of flooding. This affected 480 infrastructures. In May, 443 infrastructures were impacted, and in January, 333 infrastructures were impacted. August saw the least amount of damage, with just 137 infrastructures impacted. Table 3 summarizes the monthly infrastructure exposure to the floods.
Persistence Flooding occurred in most of the affected Infrastructure and lasted only briefly (Table 4). Specifically, 653 Infrastructures experienced two flood occurrences, and 2048 Infrastructures experienced one flood occurrence. Additionally, 533 Infrastructures were impacted by medium-persistence floods, which are 3–4 flood occurrences. Only 27 infrastructures experienced high-persistence flooding, characterized as 5–6 occurrences.
The flood recurrence (Table 5) also experienced a similar trend to the flood persistence occurrences. The highest number of infrastructures were impacted by low-recurrence events. These events only happened once. There were 2093 impacted infrastructures. 308 structures were affected by a medium-recurrence flood, which included 2 instances of flooding. In comparison, only two infrastructures were impacted by high-recurrence occurrences, which included 3 incidents of flooding.
The most common flood zone category, affecting 1812 infrastructures, was low persistence–low recurrence. Other noteworthy groups were low persistence–medium recurrence, which affected 226 infrastructure, and medium persistence–low recurrence, which impacted 351 infrastructures. No infrastructure was reported in the high persistence–high recurrence category, indicating that categories with high persistence or high recurrence were significantly less prevalent. The results are shown in Table 6.

3.3. Agricultural Losses by Crop Type

During the flood period, soybeans consistently experienced the largest flood-affected area, peaking in May (4475 ha) and remaining high in April (3409 ha) and March (2836 ha). Other crops followed similar patterns, with corn peaking in July (546 ha) and cotton in April (501 ha). Rice was most affected in April (671 ha), while wheat, although the least impacted crop overall, reached its maximum in May (35 ha). Table 7 presents the total affected area per crop for each month observed.
In the flood persistence (Table 8), the results show that the vast majority of flood-affected crop areas occurred under low-persistence conditions (1–2 flood events). For example, soybeans were impacted across approximately 6487 hectares under low persistence, while only about 27 hectares were affected under high persistence (5–6 events). A similar pattern was observed for corn, with over 1121 hectares flooded under low persistence, compared to 4 hectares under high persistence. The results indicate that most flood-related crop impacts are linked to single or short-duration flood events, rather than repeated or prolonged flooding.
Most flood-affected areas experienced low recurrence. Soybeans were most significantly impacted, covering 6725 hectares, followed by cotton at 1038 hectares and corn at 1035 hectares. However, areas affected by high recurrence had about 1 hectare of soybeans impacted, with negligible effects on other crops. Table 9 shows crop-specific affected areas (ha) under varying flood recurrence frequencies.
The flood zone category further highlighted the dominance of low persistence–low recurrence floods. This zone alone accounted for 5064 ha of soybeans, 927 ha of corn, 856 ha of cotton, and 438 ha of rice. All high persistence–high recurrence zones reported zero affected area across all crops, while medium or high categories collectively represented a very small fraction of total impact. Table 10 shows affected crop areas across flood zones.

3.4. Land Parcel Exposure and Agricultural Tenure Dynamics

The analysis of land parcel exposure and agricultural tenure dynamics across flood zones revealed distinct patterns of vulnerability based on parcel size and flood recurrence. Parcels were categorized into Small (≤48 ha), Medium (48–112 ha), and Large (>112 ha) classes, and their exposure was assessed across nine flood zones (Table 11).
In the Low Persistence–Low Recurrence zone, small parcels had the greatest exposure, with about 47,231 hectares flooded and 26,978 hectares of cultivated land affected. Medium parcels in the same zone were less affected, with 1663 hectares flooded and 1184 hectares of cultivated land impacted. Large parcels recorded 1540 hectares flooded and 917 hectares of cultivated land affected.
In zones with low persistence and medium recurrence, there was minimal exposure. Small parcels were affected by flooding over 98 hectares (ha) and 59 hectares (ha) of cultivated land. Medium parcels were flooded over 25 hectares (ha), and large parcels were flooded over 7 hectares (ha). Flood exposure in the low persistence–high recurrence zone was negligible, with less than 1 ha flooded in both small and medium parcels.
In the medium persistence–low recurrence zone, small parcels showed the greatest exposure, with 28,216 hectares flooded and 26,476 hectares of cultivated land affected. By contrast, medium parcels had 718 hectares flooded and 675 hectares of cultivated land impacted, while large parcels had only 59 hectares flooded.
In the Medium Persistence–Medium and High Recurrence zones, exposure was limited across all parcel sizes, generally to about 40 hectares or less. Cultivated land was proportionally impacted. In the high persistence–low and medium recurrence zones, exposure was minimal for all parcel sizes, with flooded areas around 8 hectares and cultivated land affected below 6 hectares.
Overall, small parcels consistently experienced the greatest cumulative flood exposure, especially in low-persistence zones. Medium and large parcels, however, were affected much less, which highlights the disproportionate vulnerability of smaller land holdings in the study area.
From Table 12, small parcels also show the highest percentage of parcel area flooded (32.650%) and percentage of parcel area that is both cultivated and flooded (18.649%). In the Low Persistence–Medium Recurrence category, flood impacts are substantially reduced across all parcel sizes. The proportion of parcels flooded remains below 1% for large and medium parcels, with slightly higher values for small parcels (0.615%). The Medium Persistence–Low Recurrence zones show a distinct increase in both proportional flooding and cultivated land intensity. Small parcels exhibit a proportion flooded of 14.788 and per-hectare intensity of 13.876%, while medium parcels show elevated values (2.985% and 2.806%, respectively). Flood exposure decreases again in the Medium Persistence–Medium Recurrence zones. Proportional flooding values remain below 0.5% for all parcel categories, and cultivated flood intensity is correspondingly low. The High Persistence–Low Recurrence zones exhibit localized but measurable impacts. Small parcels show higher proportional flooding (0.167%) and cultivated flood intensity (0.125%) compared to medium and large parcels. High Persistence–Medium Recurrence zones demonstrate proportional flooding, and cultivated flood intensity remained low across parcel sizes, though small parcels again exhibit slightly higher values. Overall, Table 12 shows that flood exposure is influenced by both flood duration and parcel size. Medium persistence zones tend to produce higher cultivated flood intensity, indicating greater sensitivity to prolonged inundation. Small parcels consistently exhibit higher proportional exposure across most flood categories, reflecting concentrated impacts within smaller land parcels.

4. Discussion

4.1. Spatial and Temporal Patterns of Structural Exposure in Relation to Flood Persistence and Recurrence

The floods in 2019 were marked by unusual spring conditions, including above-average precipitation and high snowmelt, which increased soil moisture and resulted in prevented planting throughout the Mississippi River Basin [81]. Infrastructure exposure differs significantly from urban flood effects, influenced by the spatial distribution of structures, their connections to agricultural activities, and levee systems that provide protection [82,83]. Infrastructures in rural and semi-rural areas serve as significant economic assets and indicators of local resilience. Structures such as farmhouses, storage facilities, and service infrastructure are crucial for maintaining agricultural livelihoods. Consequently, their susceptibility to flooding offers insights into how land ownership patterns and floodplain management practices influence regional vulnerability.
The monthly summary (Table 3) shows that infrastructure exposure peaked in July (480) and May (443), coinciding with the peak flood conditions along the Mississippi and Missouri Rivers during the spring and summer of 2019 [81]. This temporal pattern reflects the high-flow season and prolonged inundation, but fluctuating inundation across the lower Mississippi Basin [17,84] and aligns with hydrologic reconstructions by Muñoz et al. [13], which attributes the extensive flooding to persistent precipitation and sustained river discharge during that period. The long-lasting nature of these floods resulted in repeated and extensive inundation of structures, significantly increasing both structural damage and recovery costs. Table 4 indicates that 2048 infrastructures were impacted in low-persistence zones, while 27 infrastructures experienced significant prolonged flooding in high-persistence areas for five to six months. Such persistent flooding poses severe risks to infrastructure integrity, resulting in foundation degradation, mold growth, and permanent loss of habitability [85]. These results are consistent with the findings of Schilling et al. [86], which showed a stronger correlation between flood damage and sustained flow regimes compared to that of extreme, short-lived flooding events.
This study on the 2019 floods complements recent assessments of agricultural flood risk. Similarly to Yildirim & Demir [87], who indicated that nearly half a million acres of Iowa cropland are in a 2-year flood zone, our findings highlight significant exposure in low-recurrence areas, covering 2093 infrastructures and over 6700 hectares of soybeans. Unlike Yildirim and Demir [87], who focus on average annual losses across watersheds, we specified the critical months of May and July when infrastructure and crop exposure were highest.
Table 5 reveals that the majority of infrastructure damage (2093) was concentrated in low-recurrence areas, illustrating that most infrastructure faced flooding infrequently, occurring only once or twice during the study’s duration. In contrast, only 2 infrastructures were affected in high-recurrence zones, suggesting that repeated flooding was uncommon and localized. The combined analysis of persistence and recurrence, illustrated in Table 6, indicates a dominance of the Low Persistence–Low Recurrence category, accounting for 1812 infrastructures, which implies that the majority of flood-affected structures suffered from short-term, low-frequency flooding events. Nonetheless, a small number of infrastructures in medium- and high-persistence areas indicates significant chronic exposure areas likely intersecting with levee-protected farmlands, floodways, or conservation corridors, as noted by Reed et al. [88].
Overall, the 2019 Mississippi floods affected most infrastructure in low-persistence, low-recurrence flood zones, indicating broad short-term flood exposure in rural areas. Conversely, high-persistence flood zones, although fewer, pose significant long-term risks due to structural damage and potentially diminishing property values. These findings reinforce the importance of implementing land-use and ownership-based flood mitigation strategies that combine hydrologic persistence maps with infrastructure inventories to identify properties at risk. Such integration could enhance the allocation of flood insurance and improve infrastructure resilience planning throughout the Mississippi River Basin [86,88].

4.2. Temporal Dynamics of Flood Impacts on Crops

The monthly distribution of crops affected by floods, as shown in Table 7, shows significant temporal variations in the flood impact. Designated soybean areas were most affected by the flood, particularly in May (4475 ha), followed by corn in July (546 ha), coinciding with their critical growth stages [89], with similar occurrences in the Upper Mississippi River as reported by [90]. Wheat areas were less impacted by the flood, suggesting cultivation in less prone areas. Rice and cotton experienced moderate effects, with rice being impacted more during the spring months due to low-lying cultivation. The low-persistence floods (Table 8) cause the majority of crop damage, while medium- and high-persistence floods were less damaging overall. Short-duration floods appear more widespread, whereas prolonged floods can have severe localized impacts [91]. Soybeans and rice were most affected under these conditions, while prolonged or high-persistence floods caused severe but localized damage. Aslam et al. [92] and Colmer and Voesenek [93] confirms that short-duration floods reduce root oxygen availability, hinder photosynthesis, and trigger premature senescence, especially in crops with poor waterlogging tolerance such as corn and cotton [94]. These conditions are attributed to the loss of crops, reportedly about $600 million [89].
The flood recurrence’s impact on crop areas reveals that low recurrence floods significantly affect agricultural landscapes, particularly soybeans, corn, and cotton, with 6725 ha, 1035 ha, and 1038 ha impacted (Table 9), respectively. In contrast, areas that experience medium-to-high levels of recurrence flood frequently endure prolonged stress that negatively impacts soil fertility, structure, and nutrient cycling [95,96]. These findings align with regional assessments suggesting that frequent, low-intensity floods affect more crops than rare, extreme events [97,98]. Table 10 further highlights the compounded effects of flood persistence and recurrence on crop exposure, revealing that most affected areas were in low persistence–low recurrence zones, particularly impacting soybeans (5064 ha), cotton (856 ha), and corn (927 ha). Medium persistence–low recurrence zones also significantly affected rice (339 ha) and cotton (129 ha). High persistence–high recurrence zones had negligible impacts across all crops. Overall, the results indicate that crop exposure is strongly influenced by flood dynamics and the spatial distribution of cultivated areas. Soybeans show the highest exposed area, which is consistent with its status as the dominant crop in Mississippi, according to the Mississippi Department of Agriculture and Commerce (Mississippi Department of Agriculture and Commerce. (December 2025). Mississippi agriculture snapshot—Top 15 commodities. Retrieved from https://www.mdac.ms.gov/agency-info/mississippi-agriculture-snapshot/ (accessed on 21 April 2026)), resulting in greater spatial overlap with flood-affected zones.

4.3. Flood Zone Impact Analysis on Agricultural Land Parcels

Studies on flooding in the Mississippi River Basin highlight that the impacts on agricultural areas are shaped by intricate interactions involving the timing of floods, their characteristics, and land ownership patterns [81,99]. The analysis of the flood zone reveals significant patterns in how flooding affects parcels of different sizes and cultivated agricultural land. Results from Table 11 and Table 12 show that flood impacts differ significantly when assessed using proportional measures and cultivated flood proportions, offering a clearer understanding of relative exposure among parcel categories. The observation of a high cultivation ratio in Low persistence–Low recurrence and Medium persistence–Low recurrence, especially for small parcels, indicates that agricultural activities persist in these regions, even in the presence of flood risks. This may be attributed to the advantages of soil fertility improvements that arise from periodic flooding events. Research conducted by Dela Torre et al. [100] and Rupngam and Messiga [96] supports this notion, illustrating how farmers in areas prone to flooding often choose to continue farming activities. This decision is primarily informed by the enhanced soil fertility that flooding can provide, despite the increased risks associated with such environmental conditions.
The study indicates that Low Persistence–Low Recurrence flooding is the most impactful flood type, affecting a large area of the parcels (proportional parcel exposure) and cultivated land (cultivated land intensity). Despite its short duration and infrequent occurrence, this flood category exhibited the highest proportion of parcel area affected and the greatest cultivated flood exposure, particularly within small parcels. This suggests that even short-lived flood events can cause considerable agricultural disruption when they occur over cultivated lands. Such results are consistent with studies indicating that rare but spatially extensive floods often generate substantial agricultural damage due to their magnitude and coverage [101]. However, overall flood risk may vary due to differences in vulnerability and exposure [102,103]. Small parcels were disproportionately affected across multiple flood persistence and recurrence categories. Table 12 indicates that small parcels consistently showed higher proportions of flooded areas and greater cultivated flood intensity compared to medium and large parcels. This pattern suggests that smaller landholdings are more vulnerable to flooding. However, this vulnerability is not solely due to their greater distribution in the region, although this contributes to the total flooded area. Rather, it likely reflects a combination of spatial and structural influences. Small parcels tend to be more widely distributed across the landscape and are often located in flood-prone zones, increasing their likelihood of overlapping with inundated areas. In addition, smaller landholdings may have more limited access to protective infrastructure, such as drainage networks or levees, which further increases their susceptibility to flooding. Consequently, flooding in these areas may lead to greater relative damage to agricultural productivity. Medium parcels experienced moderate impacts, while large parcels generally exhibited lower proportional flood exposure despite accounting for larger absolute flooded areas.
Low-recurrence flooding, particularly within the low-persistence category, significantly affects agricultural land, indicating that infrequent but spatially extensive flood events pose a major concern for agricultural planning. This emphasizes the need for flood risk assessments to prioritize extreme events rather than focusing solely on frequent minor floods. Wang et al. [101] support this perspective by demonstrating that joint return period approaches incorporating flood intensity and duration better capture the risks associated with rare, high-impact agricultural flood events. In contrast, high-recurrence flooding exhibited limited proportional impacts but may represent localized and persistent inundation, potentially associated with chronic drainage issues. Although the spatial extent of such flooding is relatively small, the affected areas may require targeted mitigation strategies to address recurring agricultural disruptions.
This research aligns with findings from Andrews et al. [102] about flood risk concentration in Douglas County, Nebraska. These studies highlight that ownership structures like out-of-state and LLC holdings and distance significantly affect parcel-level flood exposure. This study expands on that by showing that small agricultural parcels face higher total area damage, especially in low-recurrence zones. Additionally, while Jahangeer et al. [104] focused on a watershed wetness index for wetland conservation from 2018 to 2023, this study uniquely integrates flood persistence with recurrence to classify structural and crop vulnerability.
This study introduces a classification of flood dynamics using persistence and recurrence, applied at agricultural, parcel, and infrastructure scales. Unlike traditional flood risk assessments that rely solely on spatial return periods, this framework highlights both the frequency and duration of flood exposure. By linking flood persistence and recurrence to land ownership and cultivated areas, the study provides a nuanced understanding of how agricultural activities persist in areas subject to medium-persistence flooding, offering insights for localized land-use planning, flood mitigation, and insurance allocation strategies that account for both temporal duration and frequency of inundation.

4.4. Limitations and Future Research

This study has several limitations related to data availability and methodological constraints. The analysis relies on multi-temporal remote sensing and LULC classifications, which are subject to uncertainties arising from cloud cover, sensor limitations, and classification errors. In particular, incomplete temporal coverage of the 2019 flood season represents a key limitation, as February and June were excluded due to insufficient cloud-free Sentinel-2 imagery. Although this may introduce minor uncertainties in the estimation of flood persistence and recurrence, the six-month dataset captures the primary flood dynamics. Future studies could improve temporal consistency by integrating Synthetic Aperture Radar (SAR) data with optical imagery.
Additionally, the study focuses on a single flood event and does not capture long-term variability across multiple years or finer temporal scales. The absence of crop yield and economic loss data also limits the ability to directly quantify productivity impacts at the farm or parcel level. Furthermore, the use of aggregated crop data does not fully account for spatial variability in topography, soil properties, drainage capacity, and farm management practices.
While the results indicate higher flood exposure among small parcels, the analysis does not explicitly distinguish whether this pattern is driven by their spatial distribution in flood-prone areas or by differences in management capacity, such as access to drainage infrastructure or levee protection. Incorporating variables such as proximity to river channels and infrastructure access would improve understanding of these dynamics.
Future research should integrate high-resolution remote sensing, hydrodynamic modeling, field-based observations, and socio-economic data to provide a more comprehensive assessment of flood impacts. In addition, incorporating crop phenology and growth-stage information would allow for a more detailed evaluation of crop-specific vulnerability to flood timing and duration.

5. Conclusions

This study demonstrates that the impacts of the 2019 Mississippi River Basin floods were driven more by spatial extent and timing of short-duration flooding than by prolonged inundation alone. Low-persistence, low-recurrence floods accounted for the majority of crop and infrastructure exposure, highlighting the significant role of infrequent but widespread flood events in shaping agricultural risk. In contrast, high-persistence flooding, although spatially limited, represents areas of chronic vulnerability with potential long-term impacts on infrastructure integrity and land usability.
The results further reveal that parcel size influences flood vulnerability, with small agricultural parcels consistently exhibiting higher proportional exposure and cultivated land intensity across flood categories. This suggests that flood risk is not only a function of hydrologic processes but is also shaped by land ownership patterns and the spatial distribution of agricultural activities.
By integrating flood persistence and recurrence with parcel-level and crop-specific analyses, this study provides a more comprehensive framework for assessing agricultural flood risk. The findings highlight the importance of incorporating both flood duration and frequency into risk assessments, particularly identifying areas exposed to large-scale, short-term flooding events that may otherwise be underestimated.
From a management perspective, the results highlight the need for targeted flood mitigation strategies that prioritize highly exposed agricultural zones, especially those dominated by smaller parcels. Integrating flood dynamics with land ownership and infrastructure data can support more effective planning, insurance allocation, and resilience-building efforts across the Mississippi River Basin.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/rs18122022/s1, Table S1. Confusion matrix for land use/land cover classification (June 2018). Table S2. Confusion matrix for land use/land cover classification (January 2019). Table S3. Confusion matrix for land use/land cover classification (March 2019). Table S4. Confusion matrix for land use/land cover classification (April 2019). Table S5. Confusion matrix for land use/land cover classification (May 2019). Table S6. Confusion matrix for land use/land cover classification (July 2019). Table S7. Confusion matrix for land use/land cover classification (August 2019). Table S8. Accuracy assessment for land use/land cover classification (June 2018). Table S9. Accuracy assessment for land use/land cover classification (January 2019). Table S10. Accuracy assessment for land use/land cover classification (March 2019). Table S11. Accuracy assessment for land use/land cover classification (April 2019). Table S12. Accuracy assessment for land use/land cover classification (May 2019). Table S13. Accuracy assessment for land use/land cover classification (July 2019). Table S14. Accuracy assessment for land use/land cover classification (August 2019). Table S15. Overall accuracy and Kappa coefficient values across the study period.

Author Contributions

Conceptualization, S.A. and J.N.M. methodology, S.A. and J.N.M.; software, J.N.M.; validation, J.N.M.; formal analysis, J.N.M.; writing—original draft preparation, J.N.M.; writing—review and editing, J.N.M. and S.A.; visualization, J.N.M.; All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

Data will be made available by request to the authors.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

References

  1. Zhang, Q.; Gu, X.; Singh, V.P.; Kong, D.; Chen, X. Spatiotemporal behavior of floods and droughts and their impacts on agriculture in China. Glob. Planet. Change 2015, 131, 63–72. [Google Scholar] [CrossRef]
  2. Balasubramanian, R.; Sreepriya, P. Soil degradation in Kerala State and a case study on socio-economic impact due to flood in its Idukki District. J. Appl. Nat. Sci. 2020, 12, 159–164. [Google Scholar] [CrossRef]
  3. de la Paix, M.J.; Lanhai, L.; Xi, C.; Ahmed, S.; Varenyam, A. Soil degradation and altered flood risk as a consequence of deforestation. Land Degrad. Dev. 2013, 24, 478–485. [Google Scholar] [CrossRef]
  4. Banerjee, L. Effects of flood on agricultural productivity in Bangladesh. Oxf. Dev. Stud. 2010, 38, 339–356. [Google Scholar] [CrossRef]
  5. Sîli, N.; Apostu, I.-M.; Faur, F. Floods and their effects on agricultural productivity. Res. J. Agric. Sci. 2020, 52, 113–122. [Google Scholar]
  6. Panichvejsunti, T.; Kuwornu, J.K.M.; Shivakoti, G.P.; Grünbühel, C.; Soni, P. Smallholder farmers’ crop combinations under different land tenure systems in Thailand: The role of flood and government policy. Land Use Policy 2018, 72, 129–137. [Google Scholar] [CrossRef]
  7. Best, J. Anthropogenic stresses on the world’s big rivers. Nat. Geosci. 2018, 12, 7–21. [Google Scholar] [CrossRef]
  8. Dunne, K.B.J.; Dee, S.G.; Reinders, J.; Muñoz, S.E.; Nittrouer, J.A. Examining the impact of emissions scenario on lower Mississippi River flood hazard projections. Environ. Res. Commun. 2022, 4, 091001. [Google Scholar] [CrossRef]
  9. Turner, R.E. Variability in the discharge of the Mississippi River and tributaries from 1817 to 2020. PLoS ONE 2022, 17, e0276513. [Google Scholar] [CrossRef]
  10. Barry, J.M. Rising Tide: The Great Mississippi Flood of 1927 and How It Changed America; Simon & Schuster: New York, NY, USA, 2007. [Google Scholar]
  11. Turner, R.E. Declining bacteria, lead, and sulphate, and rising pH and oxygen in the lower Mississippi River. Ambio 2021, 50, 1731–1738. [Google Scholar] [CrossRef]
  12. Camillo, C.A. Divine Providence: The 2011 Flood in the Mississippi River; United States Army Corps of Engineers: Omaha, NE, USA, 2012; Available online: https://digitalcommons.unl.edu/usarmyceomaha/142 (accessed on 14 February 2026).
  13. Munoz, S.E.; Giosan, L.; Therrell, M.D.; Remo, J.W.F.; Shen, Z.; Sullivan, R.M.; Wiman, C.; O’Donnell, M.; Donnelly, J.P. Climatic control of Mississippi River flood hazard amplified by river engineering. Nature 2018, 556, 95–98. [Google Scholar] [CrossRef]
  14. Su, Y.; Smith, J.A.; Villarini, G. The hydrometeorology of extreme floods in the Lower Mississippi River. J. Hydrometeorol. 2023, 24, 203–219. [Google Scholar] [CrossRef]
  15. Robinson, C.E. The 2019 Mississippi Flooding and Its Impact on Socioeconomics and Health Disparities: An Exploratory Case Study. Ph.D. Thesis, Northcentral University, San Diego, CA, USA, 2023. [Google Scholar]
  16. Criss, R.E.; Luo, M. Increasing risk and uncertainty of flooding in the Mississippi River basin. Hydrol. Process. 2017, 31, 1283–1292. [Google Scholar] [CrossRef]
  17. Pal, S.; Lee, T.R.; Clark, N.E. The 2019 Mississippi and Missouri River flooding and its impact on atmospheric boundary layer dynamics. Geophys. Res. Lett. 2020, 47, e2019GL086933. [Google Scholar] [CrossRef]
  18. Pinter, N.; Jemberie, A.A.; Remo, J.W.F.; Heine, R.A.; Ickes, B.S. Flood trends and river engineering on the Mississippi River system. Geophys. Res. Lett. 2008, 35, L23404. [Google Scholar] [CrossRef]
  19. Tripathy, S.S.; Moradkhani, H.; Moftakhari, H. A block-level categorical flood risk mapping to aid shelter location. J. Flood Risk Manag. 2025, 18, e70119. [Google Scholar] [CrossRef]
  20. Chan, J.K.H.; Liao, K.H. The normative dimensions of flood risk management: Two types of flood harm. J. Flood Risk Manag. 2022, 15, e12798. [Google Scholar] [CrossRef]
  21. Armstrong, B.N.; Cambazoglu, M.K.; Wiggert, J.D. Modeling the impact of the 2019 Bonnet Carré Spillway opening and local river flooding on the Mississippi Sound. In OCEANS 2021: San Diego–Porto; IEEE: Piscataway, NJ, USA, 2021; pp. 1–7. [Google Scholar] [CrossRef]
  22. USDA. Nation’s Wettest 12-Month Period on Record Slows Down 2019 Planting Season. U.S. Department of Agriculture. 14 June 2019. Available online: https://www.usda.gov/about-usda/news/blog/nations-wettest-12-month-period-record-slows-down-2019-planting-season (accessed on 14 February 2026).
  23. English, B.C.; Smith, S.A.; Menard, R.J.; Hughes, D.W.; Gunderson, M. Estimated economic impacts of the 2019 Midwest floods. Econ. Disasters Clim. Change 2021, 5, 431–448. [Google Scholar] [CrossRef]
  24. Newton, J. Prevent Plantings Set Record in 2019 at 20 Million Acres. Market Intel. 28 August 2019. Available online: https://www.fb.org/market-intel/prevent-plantings-set-record-in-2019-at-20-million-acres (accessed on 14 February 2026).
  25. Munawar, H.S.; Hammad, A.W.A.; Waller, S.T. Remote sensing methods for flood prediction: A review. Sensors 2022, 22, 960. [Google Scholar] [CrossRef]
  26. Zhu, W.; Cao, Z.; Luo, P.; Tang, Z.; Zhang, Y.; Hu, M.; He, B. Urban flood-related remote sensing: Research trends, gaps and opportunities. Remote Sens. 2022, 14, 5505. [Google Scholar] [CrossRef]
  27. Bates, P.D.; Quinn, N.; Sampson, C.; Smith, A.; Wing, O.; Sosa, J.; Savage, J.; Olcese, G.; Neal, J.; Schumann, G.; et al. Combined modeling of US fluvial, pluvial, and coastal flood hazard under current and future climates. Water Resour. Res. 2021, 57, e2020WR028673. [Google Scholar] [CrossRef]
  28. Callaghan, D.P.; Hughes, M.G. Assessing flood hazard changes using climate model forcing. Nat. Hazards Earth Syst. Sci. 2022, 22, 2459–2472. [Google Scholar] [CrossRef]
  29. Ebtehaj, I.; Bonakdari, H. A comprehensive comparison of CMIP5 and CMIP6 based on Canadian earth system models for long-term flood susceptibility using remote sensing and flood frequency analysis. J. Hydrol. 2023, 617, 128851. [Google Scholar] [CrossRef]
  30. Bekele, T.W.; Haile, A.T.; Trigg, M.A.; Walsh, C.L. Evaluating a new method of remote sensing for flood mapping in the urban and peri-urban areas: Applied to Addis Ababa and the Akaki catchment in Ethiopia. Nat. Hazards Res. 2022, 2, 97–110. [Google Scholar] [CrossRef]
  31. Megahed, H.A.; Abdo, A.M.; AbdelRahman, M.A.E.; Scopa, A.; Hegazy, M.N. Frequency ratio model as tools for flood susceptibility mapping in urbanized areas: A case study from Egypt. Appl. Sci. 2023, 13, 9445. [Google Scholar] [CrossRef]
  32. Sadiq, R.; Akhtar, Z.; Imran, M.; Ofli, F. Integrating remote sensing and social sensing for flood mapping. Remote Sens. Appl. Soc. Environ. 2022, 25, 100697. [Google Scholar] [CrossRef]
  33. Samanta, S.; Pal, D.K.; Palsamanta, B. Flood susceptibility analysis through remote sensing, GIS and frequency ratio model. Appl. Water Sci. 2018, 8, 66. [Google Scholar] [CrossRef]
  34. Amiri, A.; Soltani, K.; Ebtehaj, I.; Bonakdari, H. A novel machine learning tool for current and future flood susceptibility mapping by integrating remote sensing and geographic information systems. J. Hydrol. 2024, 632, 130936. [Google Scholar] [CrossRef]
  35. Avand, M.; Moradi, H.; Lasboyee, M.R. Using machine learning models, remote sensing, and GIS to investigate the effects of changing climates and land uses on flood probability. J. Hydrol. 2021, 595, 125663. [Google Scholar] [CrossRef]
  36. Farhadi, H.; Najafzadeh, M. Flood risk mapping by remote sensing data and random forest technique. Water 2021, 13, 3115. [Google Scholar] [CrossRef]
  37. Gutenson, J.L.; Follum, M.L.; Staebell, K.A.; Ondich, E.S.; Wahl, M.D. Analyzing synthetic stage-discharge rating curves and riverine flood inundation maps derived from global-scale hydrologic and hydraulic modeling. J. Flood Risk Manag. 2025, 18, e70135. [Google Scholar] [CrossRef]
  38. Huthoff, F.; Remo, J.W.F.; Pinter, N. Improving flood preparedness using hydrodynamic levee-breach and inundation modelling: Middle Mississippi River, USA. J. Flood Risk Manag. 2015, 8, 2–18. [Google Scholar] [CrossRef]
  39. Karami, M.; Abedi Koupai, J.; Gohari, S.A. Integration of SWAT, SDSM, AHP, and TOPSIS to detect flood-prone areas. Nat. Hazards 2024, 120, 6307–6325. [Google Scholar] [CrossRef]
  40. Khan, N.A.; Alzahrani, H.; Bai, S.; Hussain, M.; Tayyab, M.; Ullah, S.; Ullah, K.; Khalid, S. Flood risk assessment in the Swat River catchment through GIS-based multi-criteria decision analysis. Front. Environ. Sci. 2025, 13, 1567796. [Google Scholar] [CrossRef]
  41. Lammers, R.; Li, A.; Nag, S.; Ravindra, V. Prediction models for urban flood evolution for satellite remote sensing. J. Hydrol. 2021, 603, 127175. [Google Scholar] [CrossRef]
  42. Tedla, M.G.; Cho, Y.; Jun, K. Flood mapping from dam break due to peak inflow: A coupled rainfall–runoff and hydraulic models approach. Hydrology 2021, 8, 89. [Google Scholar] [CrossRef]
  43. Bauer-Marschallinger, B.; Cao, S.; Tupas, M.E.; Roth, F.; Navacchi, C.; Melzer, T.; Freeman, V.; Wagner, W. Satellite-Based Flood Mapping through Bayesian Inference from a Sentinel-1 SAR Datacube. Remote Sens. 2022, 14, 3673. [Google Scholar] [CrossRef]
  44. Shen, X.; Wang, D.; Mao, K.; Anagnostou, E.; Hong, Y. Inundation extent mapping by synthetic aperture radar: A review. Remote Sens. 2019, 11, 879. [Google Scholar] [CrossRef]
  45. Zhao, J.; Li, M.; Li, Y.; Matgen, P.; Chini, M. Urban Flood Mapping Using Satellite Synthetic Aperture Radar Data: A review of characteristics, approaches, and datasets. IEEE Geosci. Remote Sens. Mag. 2025, 13, 237–268. [Google Scholar] [CrossRef]
  46. Anusha, N.; Bharathi, B. Flood detection and flood mapping using multi-temporal synthetic aperture radar and optical data. Egypt. J. Remote Sens. Space Sci. 2020, 23, 207–219. [Google Scholar] [CrossRef]
  47. de Moel, H.; Jongman, B.; Kreibich, H.; Merz, B.; Penning-Rowsell, E.; Ward, P.J. Flood risk assessments at different spatial scales. Mitig. Adapt. Strateg. Glob. Change 2015, 20, 865–890. [Google Scholar] [CrossRef]
  48. Fox, S.; Agyemang, F.; Hawker, L.; Neal, J. Integrating social vulnerability into high-resolution global flood risk mapping. Nat. Commun. 2024, 15, 3155. [Google Scholar] [CrossRef] [PubMed]
  49. Samela, C.; Coluzzi, R.; Imbrenda, V.; Manfreda, S.; Lanfredi, M. Satellite flood detection integrating hydrogeomorphic and spectral indices. GIScience Remote Sens. 2022, 59, 1997–2018. [Google Scholar] [CrossRef]
  50. Saini, R.; Rawat, S.; Singh, S.R.K.; Semwal, P. Flood Mapping and Damage Analysis Using Multispectral Sentinel-2 Satellite Imagery and Machine Learning Techniques. Recent Adv. Comput. Sci. Commun. 2024, 17, E150724231944. [Google Scholar] [CrossRef]
  51. Bentivoglio, R.; Isufi, E.; Jonkman, S.N.; Taormina, R. Deep learning methods for flood mapping: A review of existing applications and future research directions. Hydrol. Earth Syst. Sci. 2022, 26, 4345–4378. [Google Scholar] [CrossRef]
  52. Chen, J.; Shi, X.; Gu, L.; Wu, G.; Su, T.; Wang, H.M.; Kim, J.S.; Zhang, L.; Xiong, L. Impacts of climate warming on global floods and their implication to current flood defense standards. J. Hydrol. 2023, 618, 129236. [Google Scholar] [CrossRef]
  53. Jones, A.; Kuehnert, J.; Fraccaro, P.; Meuriot, O.; Ishikawa, T.; Edwards, B.; Stoyanov, N.; Remy, S.L.; Weldemariam, K.; Assefa, S. AI for climate impacts: Applications in flood risk. npj Clim. Atmos. Sci. 2023, 6, 63. [Google Scholar] [CrossRef]
  54. Abu, I.O.; Ibebuchi, C.C. Risk assessment of the 2022 Nigerian flood event using remote sensing products and climate data. Remote Sens. 2025, 17, 1814. [Google Scholar] [CrossRef]
  55. Al-Omari, A.A.; Shatnawi, N.N.; Shbeeb, N.I.; Istrati, D.; Lagaros, N.D.; Abdalla, K.M. Utilizing remote sensing and GIS techniques for flood hazard mapping and risk assessment. Civ. Eng. J. 2024, 10, 1423–1436. [Google Scholar] [CrossRef]
  56. Killian, C.D.; Asquith, W.H.; Barlow, J.R.B.; Bent, G.C.; Kress, W.H.; Barlow, P.M.; Schmitz, D.W. Characterizing groundwater and surface-water interaction using hydrograph-separation techniques and groundwater-level data throughout the Mississippi Delta, USA. Hydrogeol. J. 2019, 27, 2167–2179. [Google Scholar] [CrossRef]
  57. Lo, T.H.; Pringle, H.C. A quantitative review of irrigation development in the Yazoo–Mississippi Delta from 1991 to 2020. Agronomy 2021, 11, 2548. [Google Scholar] [CrossRef]
  58. Mitch, W.A. An evaluation of the distribution of imported lithics within the Yazoo-Mississippi Delta during the Poverty Point period. Southeast. Archaeol. 2023, 42, 136–154. [Google Scholar] [CrossRef]
  59. Freshour, C.; Williams, B. Toward “Total Freedom”: Black ecologies of land, labor, and livelihoods in the Mississippi Delta. Ann. Am. Assoc. Geogr. 2023, 113, 1563–1572. [Google Scholar] [CrossRef]
  60. Allison, M.A.; Meselhe, E.A.; Kleiss, B.A.; Duffy, S.M. Impact of water loss on sustainability of the Mississippi River channel in its deltaic reach. Hydrol. Process. 2023, 37, e15004. [Google Scholar] [CrossRef]
  61. Crockett, D.; Ambinakudige, S.; Williams, B. Roots of resilience: A case study on the strength and survival of small-scale Black farmers in the Yazoo-Mississippi Delta region. J. Agric. Food Syst. Community Dev. 2025, 14, 53–70. [Google Scholar] [CrossRef]
  62. Alhassan, M.; Lawrence, C.; Richardson, S.; Pindilli, E.; U.S. Geological Survey. The Mississippi Alluvial Plain Aquifers—An Engine for Economic Activity; Fact Sheet; U.S. Geological Survey: Reston, VA, USA, 2019. [CrossRef]
  63. Massey, J.H.; Stiles, C.M.; Epting, J.W.; Powers, R.S.; Kelly, D.B.; Bowling, T.H.; Janes, C.L.; Pennington, D.A. Long-term measurements of agronomic crop irrigation made in the Mississippi Delta portion of the lower Mississippi River Valley. Irrig. Sci. 2017, 35, 297–313. [Google Scholar] [CrossRef]
  64. Rouse, J.W., Jr.; Haas, R.H.; Deering, D.W.; Schell, J.A.; Harlan, J.C. Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. In Proceedings of the Third NASA Earth Resources Technology Satellite Symposium, Washington, DC, USA, 10–14 December 1973. [Google Scholar]
  65. McFeeters, S.K. The use of the normalized difference water index (NDWI) in the delineation of open water features. Int. J. Remote Sens. 1996, 17, 1425–1432. [Google Scholar] [CrossRef]
  66. 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. [Google Scholar] [CrossRef]
  67. Zha, Y.; Gao, J.; Ni, S. Use of normalized difference built-up index in automatically mapping urban areas from TM imagery. Int. J. Remote Sens. 2003, 24, 583–594. [Google Scholar] [CrossRef]
  68. Feyisa, G.L.; Meilby, H.; Fensholt, R.; Proud, S.R. Automated water extraction index: A new technique for surface water mapping using Landsat imagery. Remote Sens. Environ. 2014, 140, 23–35. [Google Scholar] [CrossRef]
  69. Rikimaru, A.; Miyatake, S. Development of forest canopy density mapping and monitoring model using indices of vegetation, bare soil and shadow. In Proceedings of the 18th Asian Conference on Remote Sensing, Kuala Lumpur, Malaysia, 20–24 October 1997. [Google Scholar]
  70. Shen, L.; Li, C. Water body extraction from Landsat ETM+ imagery using Adaboost algorithm. In Proceedings of the 2010 18th International Conference on Geoinformatics, Beijing, China, 18–20 June 2010; pp. 1–4. [Google Scholar] [CrossRef]
  71. Sadeghi, M.; Babaeian, E.; Tuller, M.; Jones, S.B. The optical trapezoid model: A novel approach to remote sensing of soil moisture applied to Sentinel-2 and Landsat-8 observations. Remote Sens. Environ. 2017, 198, 52–68. [Google Scholar] [CrossRef]
  72. Sadeghi, M.; Mohamadzadeh, N.; Liang, L.; Bandara, U.; Caldas, M.M.; Hatch, T. A new variant of the optical trapezoid model (OPTRAM) for remote sensing of soil moisture and water bodies. Sci. Remote Sens. 2023, 8, 100105. [Google Scholar] [CrossRef]
  73. Hao, P.; Di, L.; Zhang, C.; Guo, L. Transfer learning for crop classification with Cropland Data Layer data (CDL) as training samples. Sci. Total Environ. 2020, 733, 138869. [Google Scholar] [CrossRef]
  74. Tran, K.H.; Zhang, H.K.; McMaine, J.T.; Zhang, X.; Luo, D. 10 m crop type mapping using Sentinel-2 reflectance and 30 m cropland data layer product. Int. J. Appl. Earth Obs. Geoinf. 2022, 107, 102692. [Google Scholar] [CrossRef]
  75. Chughtai, A.H.; Abbasi, H.; Karas, I.R. A review on change detection method and accuracy assessment for land use land cover. Remote Sens. Appl. Soc. Environ. 2021, 22, 100482. [Google Scholar] [CrossRef]
  76. Congalton, R.G. A review of assessing the accuracy of classifications of remotely sensed data. Remote Sens. Environ. 1991, 37, 35–46. [Google Scholar] [CrossRef]
  77. Dash, P.; Sanders, S.L.; Parajuli, P.; Ouyang, Y. Improving the accuracy of land use and land cover classification of Landsat data in an agricultural watershed. Remote Sens. 2023, 15, 4020. [Google Scholar] [CrossRef]
  78. Liu, C.; Frazier, P.; Kumar, L. Comparative assessment of the measures of thematic classification accuracy. Remote Sens. Environ. 2007, 107, 606–616. [Google Scholar] [CrossRef]
  79. Lunetta, R.S.; Fenstermaker, L.K.; Jensen, J.R.; McGwire, K.C.; Tinney, L.R. Remote sensing and geographic information system data integration: Error sources and research issues. Photogramm. Eng. Remote Sens. 1991, 57, 677–687. [Google Scholar]
  80. Dong, X.; Hu, C. A new method for describing the inundation status of floodplain wetland. Ecol. Indic. 2021, 131, 108144. [Google Scholar] [CrossRef]
  81. Kraft, L.L.; Villarini, G.; Czajkowski, J. Characterizing the 2019 Midwest flood: A hydrologic and socioeconomic perspective. Weather Clim. Soc. 2023, 15, 603–617. [Google Scholar] [CrossRef]
  82. Abedin, J.; Zou, L.; Yang, M.; Rohli, R.; Mandal, D.; Qiang, Y.; Akter, H.; Zhou, B.; Lin, B.; Cai, H. Deciphering spatial-temporal dynamics of flood exposure in the United States. Sustain. Cities Soc. 2024, 108, 105444. [Google Scholar] [CrossRef]
  83. Remo, J.W.F.; Carlson, M.; Pinter, N. Hydraulic and flood-loss modeling of levee, floodplain, and river management strategies, Middle Mississippi River, USA. Nat. Hazards 2012, 61, 551–575. [Google Scholar] [CrossRef]
  84. NOAA. 2020 Spring Flood Outlook—Update #2; National Weather Service, WFO Quad Cities: Davenport, IA, USA, 2020. Available online: https://www.weather.gov/media/dvn/Hydro/2020/2020DVNSpringFloodOutlookWebinar3_Mar122020.pdf (accessed on 14 February 2026).
  85. Aglan, H.; Ludwick, A.; Kitchens, S.; Amburgey, T.; Diehl, S.; Borazjani, H. Effect of long-term exposure and delayed drying time on moisture and mechanical integrity of flooded homes. J. Flood Risk Manag. 2014, 7, 280–288. [Google Scholar] [CrossRef]
  86. Schilling, K.E.; Anderson, E.S.; Mount, J.; Suttles, K.; Gassman, P.W.; Cerkasova, N.; White, M.J.; Arnold, J.G. Evaluation of flood metrics across the Mississippi-Atchafalaya River Basin and their relation to flood damages. PLoS ONE 2024, 19, e0307486. [Google Scholar] [CrossRef]
  87. Yildirim, E.; Demir, I. Agricultural flood vulnerability assessment and risk quantification in Iowa. Sci. Total Environ. 2022, 826, 154165. [Google Scholar] [CrossRef]
  88. Reed, T.; Mason, L.R.; Ekenga, C.C. Adapting to climate change in the Upper Mississippi River Basin: Exploring stakeholder perspectives on river system management and flood risk reduction. Environ. Health Insights 2020, 14, 1178630220984153. [Google Scholar] [CrossRef] [PubMed]
  89. Nathan, G. Agencies Tackle High Volume of Ag Damage Assessments. Mississippi State University Extension Service. 24 June 2021. Available online: https://extension.msstate.edu/news/feature-story/2021/agencies-tackle-high-volume-ag-damage-assessments (accessed on 14 February 2026).
  90. Fahie, M. Impacts of the 2019 Upper Mississippi River Flooding on Barge Movements in the Upper Midwest Region; U.S. Coast Guard, Office of Standards Evaluation and Development: Washington, DC, USA, 2019; Available online: https://www.dco.uscg.mil/Portals/9/Impacts%20of%202019%20UMR%20Flooding_Barge%20Movements_Fahie_1.pdf (accessed on 14 February 2026).
  91. Najibi, N.; Devineni, N. Recent trends in the frequency and duration of global floods. Earth Syst. Dyn. 2018, 9, 757–783. [Google Scholar] [CrossRef]
  92. Aslam, A.; Mahmood, A.; Ur-Rehman, H.; Li, C.; Liang, X.; Shao, J.; Negm, S.; Moustafa, M.; Aamer, M.; Hassan, M.U. Plant adaptation to flooding stress under changing climate conditions: Ongoing breakthroughs and future challenges. Plants 2023, 12, 3824. [Google Scholar] [CrossRef]
  93. Colmer, T.D.; Voesenek, L.A.C.J. Flooding tolerance: Suites of plant traits in variable environments. Funct. Plant Biol. 2009, 36, 665–681. [Google Scholar] [CrossRef]
  94. Mishra, B.; Chaturvedi, M.; Arya, R.; Singh, R.P.; Yadav, G. Excess soil moisture stress in maize: Physiological mechanisms and adaptive strategies. Int. J. Adv. Biochem. Res. 2024, 8, 201–211. [Google Scholar] [CrossRef]
  95. Nguyen, L.T.T.; Osanai, Y.; Anderson, I.C.; Bange, M.P.; Tissue, D.T.; Singh, B.K. Flooding and prolonged drought have differential legacy impacts on soil nitrogen cycling, microbial communities and plant productivity. Plant Soil 2018, 431, 371–387. [Google Scholar] [CrossRef]
  96. Rupngam, T.; Messiga, A.J. Unraveling the interactions between flooding dynamics and agricultural productivity in a changing climate. Sustainability 2024, 16, 6141. [Google Scholar] [CrossRef]
  97. Han, J.; Zhang, Z.; Xu, J.; Chen, Y.; Jägermeyr, J.; Cao, J.; Luo, Y.; Cheng, F.; Zhuang, H.; Wu, H.; et al. Threat of low-frequency high-intensity floods to global cropland and crop yields. Nat. Sustain. 2024, 7, 994–1006. [Google Scholar] [CrossRef]
  98. Piedra-Bonilla, E.B.; da Cunha, D.A.; Braga, M.J.; Oliveira, L.R. Extreme weather events and crop diversification: Climate change adaptation in Brazil. Mitig. Adapt. Strateg. Glob. Change 2025, 30, 5. [Google Scholar] [CrossRef]
  99. Neubert, M.; Höhnel, J.; Schinke, R. GIS-based estimation of flood damage to arable crops. AGIT J. Angew. Geoinformatik 2020, 6, 183–194. [Google Scholar] [CrossRef]
  100. Dela Torre, D.M.G.; Dela Cruz, P.K.A.; Jose, R.P.; Gatdula, N.B.; Blanco, A.C. Geospatial assessment of vulnerabilities of croplands to flooding risks: A case study of Philippine river basins. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2019, 42, 173–180. [Google Scholar] [CrossRef]
  101. Wang, Y.; Liu, G.; Guo, E.; Yun, X. Quantitative agricultural flood risk assessment using vulnerability surface and copula functions. Water 2018, 10, 1229. [Google Scholar] [CrossRef]
  102. Andrews, J.; Lee, J.; Mason, J.B.; Nam, Y.; Tang, Z. Associations between owner proximity and parcel-level flood exposure: Evidence from Douglas County, Nebraska. Int. J. Disaster Risk Reduct. 2025, 129, 105795. [Google Scholar] [CrossRef]
  103. Diaz, N.D.; Lee, Y.; Kothuis, B.L.M.; Pagán-Trinidad, I.; Jonkman, S.N.; Brody, S.D. Mapping the flood vulnerability of residential structures: Cases from The Netherlands, Puerto Rico, and the United States. Geosciences 2024, 14, 109. [Google Scholar] [CrossRef]
  104. Jahangeer, J.; Joshi, P.; Singh, R.; Lee, J.; Kapoor, A.; Tang, Z. Mapping watershed wetness dynamics and prioritizing potential agricultural land for wetland conservation programs in Nebraska. Conserv. Sci. Pract. 2025, 7, e70151. [Google Scholar] [CrossRef]
Remotesensing 18 02022 g001
Figure 1. Location of (a) the United States, (b) the state of Mississippi, (c) the Mississippi Delta region highlighting the selected counties, and (d) detailed map of the study area consisting of Washington County, Humphreys County, Holmes County, Sharkey County, Issaquena County, Warren County, and Yazoo County in the Mississippi Delta region.
Figure 1. Location of (a) the United States, (b) the state of Mississippi, (c) the Mississippi Delta region highlighting the selected counties, and (d) detailed map of the study area consisting of Washington County, Humphreys County, Holmes County, Sharkey County, Issaquena County, Warren County, and Yazoo County in the Mississippi Delta region.
Remotesensing 18 02022 g001
Remotesensing 18 02022 g002
Figure 2. Flowchart of Methodology.
Figure 2. Flowchart of Methodology.
Remotesensing 18 02022 g002
Remotesensing 18 02022 g003
Figure 3. Land use/land cover maps for the study area. (A) June 2018, (B) January 2019, (C) March 2019, (D) April 2019, (E) May 2019, (F) July 2019, (G) August 2019.
Figure 3. Land use/land cover maps for the study area. (A) June 2018, (B) January 2019, (C) March 2019, (D) April 2019, (E) May 2019, (F) July 2019, (G) August 2019.
Remotesensing 18 02022 g003
Remotesensing 18 02022 g004
Figure 4. Change Detection Binary maps. (A–F) corresponds to January, March, April, May, July, and August 2019, respectively. Each map highlights areas affected by flooding, permanent water bodies, and unchanged LULC classes.
Figure 4. Change Detection Binary maps. (A–F) corresponds to January, March, April, May, July, and August 2019, respectively. Each map highlights areas affected by flooding, permanent water bodies, and unchanged LULC classes.
Remotesensing 18 02022 g004
Remotesensing 18 02022 g005
Figure 5. Flood Extent for the Flooding Period. This figure illustrates the spatial extent of flooding for each observed month during the study period, showing the flood inundation during each flooding period.
Figure 5. Flood Extent for the Flooding Period. This figure illustrates the spatial extent of flooding for each observed month during the study period, showing the flood inundation during each flooding period.
Remotesensing 18 02022 g005
Remotesensing 18 02022 g006
Figure 6. Flood Dynamics Maps. (A) shows the Flood Persistence Map, indicating areas consistently inundated over time. (B) shows the Flood Recurrence Map, identifying locations repeatedly affected by flooding. (C) Flood Zone Map based on flood persistence and recurrence.
Figure 6. Flood Dynamics Maps. (A) shows the Flood Persistence Map, indicating areas consistently inundated over time. (B) shows the Flood Recurrence Map, identifying locations repeatedly affected by flooding. (C) Flood Zone Map based on flood persistence and recurrence.
Remotesensing 18 02022 g006
Table 1. Description of data sources.
Table 1. Description of data sources.
Data Type DescriptionSource
DEMUSGS 3DEP 10 m National Map Seamless DEM (10 m)https://developers.google.com/earth-engine/datasets/catalog/USGS_3DEP_10m_collection (Accessed on 20 October 2025)
Cropland Data LayerCrop-specific land cover data layer (30 m)https://developers.google.com/earth-engine/datasets/catalog/USDA_NASS_CDL (Accessed on 20 October 2025)
Sentinel 2Harmonized Sentinel-2 MSI: Multispectral Instrument, Level-2A (SR) (10 m, 20, 60 m)https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR_HARMONIZED (Accessed on 20 October 2025)
Building FootprintBuilding footprint polygons of Mississippi structureshttps://maris.mississippi.edu/HTML/DATA/data_Cadastral/BuildingFootprints.html#gsc.tab=0. (Accessed on 25 October 2025)
Land Parcel Cadastral frameworkhttps://maris.mississippi.edu/HTML/DATA/data_Cadastral/CadastralFramework.html#gsc.tab=0 (Accessed on 25 October 2025)
Study Area ShapefileThe United States Census Bureau TIGER dataset for the boundaries of the primary legal divisions of the US states counties.https://developers.google.com/earth-engine/datasets/catalog/TIGER_2018_Counties (Accessed on 19 October 2025)
Table 2. Description of land cover classes.
Table 2. Description of land cover classes.
Land Cover TypeDescriptionCDL Code
Water BodiesMississippi River, Fishponds, open water, lakes, reservoirs, and streams92,111
Agricultural AreasCrop land areas, such as Corn, Rice, Soybeans, Cotton, Wheat fields, and Vegetables, etc.1,2,3,4,5,22,23,24
Forest AreasDense tree covers, including Forest areas like the Deciduous, Evergreen, and Mixed Forest areas.141,142,143
Shrubland and GrasslandAreas dominated by shrubs, grasses, or herbaceous vegetation152,176
Built-up AreasUrban and rural settlements, infrastructure, roads, pavements, and impervious surfaces 121,122,123,124
Barren LandsAny open land areas with minimal or no vegetation, such as exposed soil, sand, rock, or degraded land.131
Source: USDA-NASS (https://www.nass.usda.gov/Research_and_Science/Cropland/Viewer/index.php (Accessed on 20 October 2025)).
Table 3. Monthly summary of infrastructure exposure to flooding events.
Table 3. Monthly summary of infrastructure exposure to flooding events.
MonthNumber of Infrastructures Affected
January333
March267
April192
May443
July480
August137
Table 4. Number of infrastructures affected in persistent flooding areas.
Table 4. Number of infrastructures affected in persistent flooding areas.
PersistenceNumber of Infrastructure AffectedPersistence Category
12048Low Persistence
2653Low Persistence
3337Medium Persistence
4196Medium Persistence
525High Persistence
62High Persistence
Table 5. Number of infrastructures affected in recurrent flooding areas.
Table 5. Number of infrastructures affected in recurrent flooding areas.
RecurrenceNumber of Infrastructure AffectedRecurrence Category
12093Low Recurrence
2308Medium Recurrence
32High Recurrence
Table 6. Number of infrastructures affected in flood zone areas.
Table 6. Number of infrastructures affected in flood zone areas.
Flood Zone CategoryNumber of Infrastructures Affected
Low Persistence–Low Recurrence1812
Low Persistence–Medium Recurrence226
Low Persistence–High Recurrence3
Medium Persistence–Low Recurrence351
Medium Persistence–Medium Recurrence127
Medium Persistence–High Recurrence2
High Persistence–Low Recurrence13
High Persistence–Medium Recurrence1
High Persistence–High Recurrence0
Table 7. Affected crop areas (ha) across different months.
Table 7. Affected crop areas (ha) across different months.
CropsJanuaryMarchAprilMayJulyAugust
Corn16822819023754666
Wheat2.070242035301
Soybeans1454283634094475832205
Rice21349567163835482
Cotton19044650124626717
Table 8. Crop area affected (ha) by varying levels of flood persistence.
Table 8. Crop area affected (ha) by varying levels of flood persistence.
Persistence CategoryFlood PersistenceCornCottonRiceSoybeansWheat
Low Persistence11015790385486315
Low Persistence2106201161162312
Medium Persistence34113321313744
Medium Persistence4141420521013
High Persistence53247251
High Persistence61<1820
Table 9. Crop area (ha) affected by flood recurrence levels.
Table 9. Crop area (ha) affected by flood recurrence levels.
RecurrenceRecurrence CategoryCornCottonRiceSoybeansWheat
1Low Recurrence10351038814672542
2Medium Recurrence2229252342
3High Recurrence<1<1<110
Table 10. Affected crop areas (ha) across combined flood zone categories.
Table 10. Affected crop areas (ha) across combined flood zone categories.
Flood ZoneCornCottonRiceSoybeansWheat
Low Persistence–Low Recurrence927856438506424
Low Persistence–Medium Recurrence13258181<1
Low Persistence–High Recurrence<10<1<10
Medium Persistence–Low Recurrence46129339142116
Medium Persistence–Medium Recurrence7315501
Medium Persistence–High Recurrence0<1<110
High Persistence–Low Recurrence2<11511<1
High Persistence–Medium Recurrence<1<1<110
High Persistence–High Recurrence00000
Table 11. Flood exposure and agricultural tenure dynamics across the flood zones.
Table 11. Flood exposure and agricultural tenure dynamics across the flood zones.
Flood Zone CategoryParcel CategoryTotal Parcel Flooded (ha)Cultivated Land Flooded (ha)
Low Persistence–Low RecurrenceLarge1540918
Low Persistence–Low RecurrenceMedium16631184
Low Persistence–Low RecurrenceSmall47,23126,978
Low Persistence–Medium RecurrenceLarge75
Low Persistence–Medium RecurrenceMedium2521
Low Persistence–Medium RecurrenceSmall9859
Low Persistence–High RecurrenceMedium<1<1
Low Persistence–High RecurrenceSmall<1<1
Medium Persistence–Low RecurrenceLarge5922
Medium Persistence–Low RecurrenceMedium718675
Medium Persistence–Low RecurrenceSmall28,21626,476
Medium Persistence–Medium RecurrenceLarge53
Medium Persistence–Medium RecurrenceMedium74
Medium Persistence–Medium RecurrenceSmall4032
Medium Persistence–High RecurrenceMedium<1<1
Medium Persistence–High RecurrenceSmall<1<1
High Persistence–Low RecurrenceLarge<1<1
High Persistence–Low RecurrenceMedium32
High Persistence–Low RecurrenceSmall86
High Persistence–Medium RecurrenceLarge<1<1
High Persistence–Medium RecurrenceMedium<1<1
High Persistence–Medium RecurrenceSmall<1<1
Note: Estimated areas less than 1 hectare are denoted as “<1”. Small: ≤48 ha, Medium: 48–112 ha, Large: >112 ha.
Table 12. Proportional parcel flood exposure and cultivated land intensity by parcel size and flood persistence–recurrence class.
Table 12. Proportional parcel flood exposure and cultivated land intensity by parcel size and flood persistence–recurrence class.
Flood Zone CategoryTotal Parcel (ha)Proportion of Parcel Flooded (%)Cultivated Flood Exposure (%)
Low Persistence–Low Recurrence259,016.6350.5950.354
Low Persistence–Low Recurrence161,682.4861.0290.732
Low Persistence–Low Recurrence144,658.52132.65018.649
Low Persistence–Medium Recurrence27,393.4490.0260.018
Low Persistence–Medium Recurrence16,156.4290.1550.130
Low Persistence–Medium Recurrence15,940.2730.6150.370
Low Persistence–High Recurrence133.0500.0080.004
Low Persistence–High Recurrence0.4591.8310.000
Medium Persistence–Low Recurrence47,507.5590.1240.046
Medium Persistence–Low Recurrence24,055.2992.9852.806
Medium Persistence–Low Recurrence190,807.69314.78813.876
Medium Persistence–Medium Recurrence20,176.5000.0250.015
Medium Persistence–Medium Recurrence12,190.7760.0570.033
Medium Persistence–Medium Recurrence10,316.8420.3880.310
Medium Persistence–High Recurrence129.4660.0110.005
Medium Persistence–High Recurrence84.6650.1210.033
High Persistence–Low Recurrence9496.9550.0090.005
High Persistence–Low Recurrence5692.6350.0530.035
High Persistence–Low Recurrence4794.1320.1670.125
High Persistence–Medium Recurrence426.5020.0070.006
High Persistence–Medium Recurrence267.3740.0070.003
High Persistence–Medium Recurrence233.8220.1350.077
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.

Share and Cite

MDPI and ACS Style

Marfo, J.N.; Ambinakudige, S. Unequal Burdens: Land Tenure and Agricultural Losses in the 2019 Lower Mississippi River Floods. Remote Sens. 2026, 18, 2022. https://doi.org/10.3390/rs18122022

AMA Style

Marfo JN, Ambinakudige S. Unequal Burdens: Land Tenure and Agricultural Losses in the 2019 Lower Mississippi River Floods. Remote Sensing. 2026; 18(12):2022. https://doi.org/10.3390/rs18122022

Chicago/Turabian Style

Marfo, Jephthah Nimoh, and Shrinidhi Ambinakudige. 2026. "Unequal Burdens: Land Tenure and Agricultural Losses in the 2019 Lower Mississippi River Floods" Remote Sensing 18, no. 12: 2022. https://doi.org/10.3390/rs18122022

APA Style

Marfo, J. N., & Ambinakudige, S. (2026). Unequal Burdens: Land Tenure and Agricultural Losses in the 2019 Lower Mississippi River Floods. Remote Sensing, 18(12), 2022. https://doi.org/10.3390/rs18122022

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop