Deltaic islands are distinct hydro-environmental zones with global significance in food security, biodiversity conservation, and fishery industry. These islands are the fundamental building blocks of many river deltas. However, deltaic islands are facing severe challenges due to intensive anthropogenic activities, sea level rise, and climate change. In this study, dynamic changes of deltaic islands in Wax Lake Delta (WLD) and Atchafalaya Delta (AD), part of the Atchafalaya River Delta Complex (ARDC) in Louisiana, USA, were quantified based on remote sensing images from 1991 to 2019 through a machine learning method. Results indicate a significant increase in deltaic islands area for the whole ARDC at a rate of 1.29 km2/yr, with local expansion rates of 0.60 km2/yr for WLD and 0.69 km2/yr for AD. All three parts of the WLD naturally prograded seaward, with the western part (WP) and central part (CP) expanding southwestward to the sea, while the eastern part (EP) prograding southeastwards. Differently from WLD, the three parts of AD irregularly expanded seaward under the impacts of anthropogenic activities. The WP and CP of the AD expanded respectively northwestwards and southwestwards, while the EP remained stable. Different drivers dominate the growth of deltaic islands in the WLD and AD. Specifically, fluvial suspended sediment discharge and peak flow events were responsible for the shift in the spatial evolution of WLD, while dredging and sediment disposal contributed to the expansion of AD. Tropical storms with different intensity and landing locations caused short-term deltaic island erosion or expansion. Tropical storms mainly generated erosion on the deltaic islands of the WLD, while causing transient erosion or siltation on the deltaic islands of the AD. In addition, high-intensity hurricanes that made landfall east of the deltas caused more erosion in the AD. Finally, sea level rise, at the current rate of 8.17 mm/yr, will not pose a threat to the deltaic island of WLD, while the eastern part of AD may be at risk of drowning. This study recognizes the complexity of factors influencing the growth of deltaic islands, suggesting that quantitative studies on the deltaic island extent are of critical for the restoration and sustainable management of the Mississippi River Delta and other deltas around the world.
more »
« less
Integrating Satellite Imagery and Ground-Based Measurements with a Machine Learning Model for Monitoring Lake Dynamics over a Semi-Arid Region
The long-term variability of lacustrine dynamics is influenced by hydro-climatological factors that affect the depth and spatial extent of water bodies. The primary objective of this study is to delineate lake area extent, utilizing a machine learning approach, and to examine the impact of these hydro-climatological factors on lake dynamics. In situ and remote sensing observations were employed to identify the predominant explanatory pathways for assessing the fluctuations in lake area. The Great Salt Lake (GSL) and Lake Chad (LC) were chosen as study sites due to their semi-arid regional settings, enabling the testing of the proposed approach. The random forest (RF) supervised classification algorithm was applied to estimate the lake area extent using Landsat imagery that was acquired between 1999 and 2021. The long-term lake dynamics were evaluated using remotely sensed evapotranspiration data that were derived from MODIS, precipitation data that were sourced from CHIRPS, and in situ water level measurements. The findings revealed a marked decline in the GSL area extent, exceeding 50% between 1999 and 2021, whereas LC exhibited greater fluctuations with a comparatively lower decrease in its area extent, which was approximately 30% during the same period. The framework that is presented in this study demonstrates the reliability of remote sensing data and machine learning methodologies for monitoring lacustrine dynamics. Furthermore, it provides valuable insights for decision makers and water resource managers in assessing the temporal variability of lake dynamics.
more »
« less
- Award ID(s):
- 1829764
- NSF-PAR ID:
- 10451791
- Date Published:
- Journal Name:
- Hydrology
- Volume:
- 10
- Issue:
- 4
- ISSN:
- 2306-5338
- Page Range / eLocation ID:
- 78
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)Abstract. The empirical attribution of hydrologic change presents a unique data availability challenge in terms of establishing baseline prior conditions, as one cannot go back in time to retrospectively collect the necessary data. Although global remote sensing data can alleviate this challenge, most satellite missions are too recent to capture changes that happened long ago enough to provide sufficient observations for adequate statistical inference. In that context, the 4 decades of continuous global high-resolution monitoring enabled by the Landsat missions are an unrivaled source of information. However, constructing a time series of land cover observation across Landsat missions remains a significant challenge because cloud masking and inconsistent image quality complicate the automatized interpretation of optical imagery. Focusing on the monitoring of lake water extent, we present an automatized gap-filling approach to infer the class (wet or dry) of pixels masked by clouds or sensing errors. The classification outcome of unmasked pixels is compiled across images taken on different dates to estimate the inundation frequency of each pixel, based on the assumption that different pixels are masked at different times. The inundation frequency is then used to infer the inundation status of masked pixels on individual images through supervised classification. Applied to a variety of global lakes with substantial long term or seasonal fluctuations, the approach successfully captured water extent variations obtained from in situ gauges (where applicable), or from other Landsat missions during overlapping time periods. Although sensitive to classification errors in the input imagery, the gap-filling algorithm is straightforward to implement on Google's Earth Engine platform and stands as a scalable approach to reliably monitor, and ultimately attribute, historical changes in water bodies.more » « less
-
more » « less
Abstract The Consistent Artificial Intelligence (AI)-based Soil Moisture (CASM) dataset is a global, consistent, and long-term, remote sensing soil moisture (SM) dataset created using machine learning. It is based on the NASA Soil Moisture Active Passive (SMAP) satellite mission SM data and is aimed at extrapolating SMAP-like quality SM back in time using previous satellite microwave platforms. CASM represents SM in the top soil layer, and it is defined on a global 25 km EASE-2 grid and for 2002–2020 with a 3-day temporal resolution. The seasonal cycle is removed for the neural network training to ensure its skill is targeted at predicting SM extremes. CASM comparison to 367 global
in-situ SM monitoring sites shows a SMAP-like median correlation of 0.66. Additionally, the SM product uncertainty was assessed, and both aleatoric and epistemic uncertainties were estimated and included in the dataset. CASM dataset can be used to study a wide range of hydrological, carbon cycle, and energy processes since only a consistent long-term dataset allows assessing changes in water availability and water stress. -
Ocean colour is recognised as an Essential Climate Variable (ECV) by the Global Climate Observing System (GCOS); and spectrally-resolved water-leaving radiances (or remote-sensing reflectances) in the visible domain, and chlorophyll-a concentration are identified as required ECV products. Time series of the products at the global scale and at high spatial resolution, derived from ocean-colour data, are key to studying the dynamics of phytoplankton at seasonal and inter-annual scales; their role in marine biogeochemistry; the global carbon cycle; the modulation of how phytoplankton distribute solar-induced heat in the upper layers of the ocean; and the response of the marine ecosystem to climate variability and change. However, generating a long time series of these products from ocean-colour data is not a trivial task: algorithms that are best suited for climate studies have to be selected from a number that are available for atmospheric correction of the satellite signal and for retrieval of chlorophyll-a concentration; since satellites have a finite life span, data from multiple sensors have to be merged to create a single time series, and any uncorrected inter-sensor biases could introduce artefacts in the series, e.g., different sensors monitor radiances at different wavebands such that producing a consistent time series of reflectances is not straightforward. Another requirement is that the products have to be validated against in situ observations. Furthermore, the uncertainties in the products have to be quantified, ideally on a pixel-by-pixel basis, to facilitate applications and interpretations that are consistent with the quality of the data. This paper outlines an approach that was adopted for generating an ocean-colour time series for climate studies, using data from the MERIS (MEdium spectral Resolution Imaging Spectrometer) sensor of the European Space Agency; the SeaWiFS (Sea-viewing Wide-Field-of-view Sensor) and MODIS-Aqua (Moderate-resolution Imaging Spectroradiometer-Aqua) sensors from the National Aeronautics and Space Administration (USA); and VIIRS (Visible and Infrared Imaging Radiometer Suite) from the National Oceanic and Atmospheric Administration (USA). The time series now covers the period from late 1997 to end of 2018. To ensure that the products meet, as well as possible, the requirements of the user community, marine-ecosystem modellers, and remote-sensing scientists were consulted at the outset on their immediate and longer-term requirements as well as on their expectations of ocean-colour data for use in climate research. Taking the user requirements into account, a series of objective criteria were established, against which available algorithms for processing ocean-colour data were evaluated and ranked. The algorithms that performed best with respect to the climate user requirements were selected to process data from the satellite sensors. Remote-sensing reflectance data from MODIS-Aqua, MERIS, and VIIRS were band-shifted to match the wavebands of SeaWiFS. Overlapping data were used to correct for mean biases between sensors at every pixel. The remote-sensing reflectance data derived from the sensors were merged, and the selected in-water algorithm was applied to the merged data to generate maps of chlorophyll concentration, inherent optical properties at SeaWiFS wavelengths, and the diffuse attenuation coefficient at 490 nm. The merged products were validated against in situ observations. The uncertainties established on the basis of comparisons with in situ data were combined with an optical classification of the remote-sensing reflectance data using a fuzzy-logic approach, and were used to generate uncertainties (root mean square difference and bias) for each product at each pixel.more » « less
-
The phenology of critical biological events in aquatic ecosystems are rapidly shifting due to climate change. Growing variability in phenological cues can increase the likelihood of trophic mismatches, causing recruitment failures in commercially, culturally, and recreationally important fisheries. We tested for changes in spawning phenology of regionally important walleye (Sander vitreus) populations in 194 Midwest US lakes in Minnesota, Michigan, and Wisconsin spanning 1939-2019 to investigate factors influencing walleye phenological responses to climate change and associated climate variability, including ice-off timing, lake physical characteristics, and population stocking history. Data from Wisconsin and Michigan lakes (185 and 5 out of 194 total lakes, respectively) were collected by the Wisconsin Department of Natural Resources (WDNR) and the Great Lakes Indian Fish and Wildlife Commission (GLIFWC) through standardized spring walleye mark-recapture surveys and spring tribal harvest season records. Standardized spring mark-recapture population estimates are performed shortly after ice-off, where following a marking event, a subsequent recapture sampling event is conducted using nighttime electrofishing (typically AC – WDNR, pulsed-DC – GLIFWC) of the entire shoreline including islands for small lakes and index stations for large lakes (Hansen et al. 2015) that is timed to coincide with peak walleye spawning activity (G. Hatzenbeler, WDNR, personal communication; M. Luehring, GLIFWC, personal communication; Beard et al. 1997). Data for four additional Minnesota lakes were collected by the Minnesota Department of Natural Resources (MNDNR) beginning in 1939 during annual collections of walleye eggs and broodstock (Schneider et al. 2010), where date of peak egg take was used to index peak spawning activity. For lakes where spawning location did not match the lake for which the ice-off data was collected, the spawning location either flowed into (Pike River) or was within 50 km of a lake where ice-off data were available (Pine River) and these ice-off data were used. Following the affirmation of off-reservation Ojibwe tribal fishing rights in the Ceded Territories of Wisconsin and the Upper Peninsula of Michigan in 1987, tribal spearfishers have targeted walleye during spring spawning (Mrnak et al. 2018). Nightly harvests are recorded as part of a compulsory creel survey (US Department of the Interior 1991). Using these records, we calculated the date of peak spawning activity in a given lake-year as the day of maximum tribal harvest. Although we were unable to account for varying effort in these data, a preliminary analysis comparing spawning dates estimated using tribal harvest to those determined from standardized agency surveys in the same lake and year showed that they were highly correlated (Pearson’s correlation: r = 0.91, P < 0.001). For lakes that had walleye spawning data from both agency surveys and tribal harvest, we used the data source with the greatest number of observation years. Ice-off phenology data was collected from two sources – either observed from the Global Lake and River Ice Phenology database (Benson et al. 2000)t, or modeled from a USGS region-wide machine-learning model which used North American Land Data Assimilation System (NLDAS) meteorological inputs combined with lake characteristics (lake position, clarity, size, depth, hypsography, etc.) to predict daily water column temperatures from 1979 - 2022, from which ice-off dates could be derived (https://www.sciencebase.gov/catalog/item/6206d3c2d34ec05caca53071; see Corson-Dosch et al. 2023 for details). Modeled data for our study lakes (see (Read et al. 2021) for modeling details), which performed well in reflecting ice phenology when compared to observed data (i.e., highly significant correlation between observed and modeled ice-off dates when both were available; r = 0.71, p < 0.001). Lake surface area (ha), latitude, and maximum depth (m) were acquired from agency databases and lake reports. Lake class was based on a WDNR lakes classification system (Rypel et al. 2019) that categorized lakes based on temperature, water clarity, depth, and fish community. Walleye stocking history was defined using the walleye stocking classification system developed by the Wisconsin Technical Working Group (see also Sass et al. 2021), which categorized lakes based on relative contributions of naturally-produced and stocked fish to adult recruitment by relying heavily on historic records of age-0 and age-1 catch rates and stocking histories. Wisconsin lakes were divided into three groups: natural recruitment (NR), a combination of stocking and natural recruitment (C-ST), and stocked only (ST). Walleye natural recruitment was indexed as age-0 walleye CPE (number of age-0 walleye captured per km of shoreline electrofished) from WDNR and GLIFWC fall electrofishing surveys (see Hansen et al. 2015 for details). We excluded lake-years where stocking of age-0 fish occurred before age-0 surveys to only include measurements of naturally-reproduced fish.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3390/hydrology10040078
