Comparing Landsat Dynamic Surface Water Extent to alternative methods of measuring inundation in developing waterbird habitats

John B. Taylor; Jeffery D. Sullivan; Claire S. Teitelbaum; Jan G. Reese; Diann J. Prosser

doi:10.1016/j.rsase.2022.100845

Home/Archive/USGS

USGS · 70237388

Comparing Landsat Dynamic Surface Water Extent to alternative methods of measuring inundation in developing waterbird habitats

John B. Taylor·Jeffery D. Sullivan·Claire S. Teitelbaum·Jan G. Reese·Diann J. Prosser

View original source ↗Export metadata ↓

Abstract

This study investigates the applicability of the Landsat Dynamic Surface Water Extent (DSWE) science product for waterbird habitat modeling in multiple non-canopied habitat types. We compare surface water distribution estimates derived from DSWE to two site-specific survey methods: visual surveys and digitized aerial imagery. These site-specific surveys were conducted on Poplar Island, a restoration island project in the Chesapeake Bay, USA. Visual surveys were collected bimonthly from 2006 – 2013, and digitized aerial imagery was collected annually from 2006 – 2015. As a restoration island, Poplar Island presents a unique opportunity to analyze DSWE in a rapidly changing site. We structure our analysis based on the procedural development of individual sub-island cells developed from unconsolidated dredge material into fully restored wetlands that have independent hydrologic connection to the surrounding bay. Each development status is analyzed using our three DSWE classifications: Open Water (OW), a conservative estimate; Wetland Inclusive (WI), an aggressive estimate; and Development Dependent (DD), a landcover adaptive estimate. The OW classification consistently underestimates surface water coverage especially in the more complex, fully developed cells. The WI classification is better able to capture the tidal channels in these cells, but marginally overestimates surface water coverage in more sparsely vegetated cells. The DD classification does not significantly improve upon the estimations of the WI classification. Our data indicate that DSWE can be a capable alternative to our site-specific survey methods. However, the product is limited by Landsat’s 30 m spatial resolution, especially in more structurally complex wetlands. A recommended classification method for characterizing waterbird habitats would depend on the goals and targeted scale of analysis, for which DSWE may be a viable option.

Explore related subjects

Remote Sensing Applications: Society and Environment Article

Source-reported bounding extent: 38.741766321754575° to 38.797577240505625° latitude; -76.39291763305664° to -76.35772705078125° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

John B. Taylor, Jeffery D. Sullivan, Claire S. Teitelbaum, Jan G. Reese, Diann J. Prosser. 2022. Comparing Landsat Dynamic Surface Water Extent to alternative methods of measuring inundation in developing waterbird habitats. https://doi.org/10.1016/j.rsase.2022.100845

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

Regreening, restoring, and reconnecting a southwestern wetland ecosystem – the Zeedyk wetland

Alluvial wetland ecosystems are vital as biodiversity hotspots but are increasingly threatened by anthropogenic stressors and drought. These pressures are especially acute in arid and semi-arid regions, where eco-hydrologic connectivity is fragile and recovery is slow. This study quantifies the efficacy of nature-based solutions, particularly the ‘Zeedyk approach,’ which employs low-tech Natural Infrastructure in Dryland Streams (NIDS)—including rock detention structures—to slow surface water, raise groundwater tables, and restore wetland function at a spring-fed wetland in Cebolla Canyon, New Mexico, U.S.A. Our results depict a Restoration Feedback Loop that captures stages of change from a healthy wetland in 1935, altered by 20th-century agriculture and grazing, to the re-establishment of the historical flow regime by 2024 documented through an 89-year archive of aerial imagery (1935–2024). By the end of our study period, the Spring-Fed Wetland had expanded by roughly 229% of the original 1935 area, to 4.13 ha. Using 40 years of satellite data, we assess changes in vegetation and hydrology with remote sensing indices. Spatial and temporal analyses reveal significant increases in vegetation greenness and wetness, particularly in an Expanded Wetland subregion, which exhibited ∼3.5x higher wetness and ∼1.5x higher greenness trends compared to adjacent areas. Monthly metrics highlight seasonal variability, with increases in greenness linked to monsoonal rainfall and lateral water redistribution, indicating that restoration impacts extend beyond the primary wetland. This study demonstrates the utility of cloud-based platforms like Google Earth Engine and USGS EarthExplorer for long-term monitoring of wetland restoration, while quantifying the efficacy of the ‘Zeedyk approach’ and demonstrating its potential as a scalable method to restore and conserve wetland meadows in other arid and semi-arid landscapes.

New Mexico↗

USGS2026

A multidecadal metric of coastal wetland stability: Development of the Landsat-based UnVegetated-to-Vegetated Ratio (UVVR) version 2 spanning 1985-2022

The UnVegetated-to-Vegetated Ratio (UVVR) is a well-established metric of wetland condition, identifying stable wetlands with high vegetation cover and low vegetation cover wetlands prone to loss, informing research and management efforts in the coastal zone. In this study, we document the development and release of version 2 of UVVR (UVVR v2 ) for the coastal conterminous United States. UVVR v2 was derived from Landsat Collection-2 satellite imagery providing increased image consistency and geolocation accuracy. We describe several UVVR v2 updates including expanding the product record from 2014–2018 to 1985–2022 and improved calibration of spectral indices used to estimate fractional cover. We validated UVVR v2 and UVVR v1 against the MODIS Vegetation Continuous Field (VCF) products and demonstrate that UVVR v2 agrees well with VCF vegetation cover across the Atlantic, Gulf, and Pacific Coasts, with R 2 > 0.82 and root mean square errors (RMSEs) < 0.18 over five years. Relative to the UVVR v1 , UVVR v2 showed increases in R 2 > 0.29 and reductions in RMSEs > 0.12. Moving beyond metric development, we present several examples of UVVR v2 applications including multidecadal vegetation cover trend maps for various coastal wetland systems. Our findings demonstrate an increased ability to characterize wetland condition with Landsat Collection-2 imagery, and indicate that upcoming missions (e.g., Landsat Next) could result in further characterization improvements. Overall, UVVR v2 helps advance coastal wetlands monitoring and management efforts due to increased accuracy and longer period-of-record.

conterminous United States↗

USGS2026

Comparative crop yield forecasting using satellite-derived biophysical and agro-climatic predictors in Sub-Saharan Africa

Timely and accurate crop yield forecasting is central to food security early warning systems, particularly in climate-vulnerable regions. While operational forecasting frameworks commonly rely on precipitation and vegetation indices such as NDVI, their ability to provide actionable lead time remains limited. Here, we evaluate the added value of satellite-derived biophysical Essential Climate Variables (ECVs): Leaf Area Index (LAI) and Fraction of Photosynthetically Active Radiation (FAPAR), for forecasting millet yield in Burkina Faso (BF) and maize yield in South Africa (ZA) and Malawi (MW). Using Random Forest models, we quantify forecast skill across the growing season at both national and sub-national scales. Results show that LAI and FAPAR provide effective forecast lead times of approximately 4 months in BF, 2 months in ZA, and up to 6 months in MW relative to harvest. At peak performance, Mean Absolute Percentage Error (MAPE) reaches 19.8% (LAI) and 23.8% (FAPAR) in BF, 12.0% and 9.8% in ZA, and 21.8% and 20.8% in MW, respectively. Across countries, biophysical parameters often outperform NDVI and precipitation, particularly in arid and semi-arid regions. At the sub-national level, LAI and FAPAR enable classification of administrative units into high and moderate-skill forecast units, revealing strong spatial heterogeneity linked to crop dominance. However, forecast skill declines where the target crop is not the dominant type, highlighting an important limitation for operational deployment. Overall, the findings suggest that satellite-derived biophysical parameters can provide earlier and more spatially resolved yield signals than commonly used predictors, with potential to improve the timeliness and effectiveness of food security early warning systems.

Remote Sensing Applications: Society and Environme↗