Unveiling Significant Shoreline Changes in Lake Michigan After a Record-Setting Water Level Increase using High-Resolution Satellite Images
Authors:
Hazem U. Abdelhady,
Cary D. Troy,
Longhuan Zhu,
Pengfei Xue,
Guy Meadows,
Chin H. Wu
Abstract:
In this paper, high-resolution multispectral satellite images were used to uncover a remarkable shoreline transformation in Lake Michigan coastal areas, driven by a record-setting increase in the water level between 2013 and 2020. Shoreline change analyses were conducted for eleven different natural beaches around the lake, unveiling significant variations of shoreline retreat despite being affect…
▽ More
In this paper, high-resolution multispectral satellite images were used to uncover a remarkable shoreline transformation in Lake Michigan coastal areas, driven by a record-setting increase in the water level between 2013 and 2020. Shoreline change analyses were conducted for eleven different natural beaches around the lake, unveiling significant variations of shoreline retreat despite being affected by the same water level increase. The average observed shoreline retreats between 2013 and 2020 for the beaches ranged between 20 m and 62 m. When the passive inundation was excluded, the estimated morphological changes were found to differ significantly from site to site, with some locations experiencing minimal changes, while others encountered considerable morphological changes of up to 38m. The examination of the correlation between the morphological changes and ten hydrodynamic and morphological factors revealed strong correlations with the offshore slopes and beach width, with steeply slo**, wide beaches experiencing more erosion. Notably, wave power, longshore sediment transport divergence, and the number of storms exhibited moderate correlation with the observed morphological changes. The results of the shoreline changes and correlation analysis offer valuable insights into the varied effects of increased water levels on Lake Michigan beaches, including erosion and passive inundation, while shedding light on the key factors driving shoreline erosion in this context. These insights can help decision and policymakers in making informed choices regarding the protection and management of Lake Michigan coastal areas, particularly in anticipation of future incidents of water level increase.
△ Less
Submitted 25 September, 2023;
originally announced September 2023.
A Machine Learning Framework for Extending Wave Height Time Series Using Historical Wind Records
Authors:
Hazem U. Abdelhady,
Cary D. Troy
Abstract:
This study presents a novel machine learning-based (ML) framework that utilizes the ConvLSTM-1D model to hindcast or forecast wave heights at coastal locations using a nonuniform array of wind observations. This approach was applied to Lake Michigan to perform a 70-year ice-free hindcast of waves near Chicago, IL (USA). The Wave Information System model (WIS) served as the training, validation, an…
▽ More
This study presents a novel machine learning-based (ML) framework that utilizes the ConvLSTM-1D model to hindcast or forecast wave heights at coastal locations using a nonuniform array of wind observations. This approach was applied to Lake Michigan to perform a 70-year ice-free hindcast of waves near Chicago, IL (USA). The Wave Information System model (WIS) served as the training, validation, and testing dataset for the ML model. Ensemble learning-optimized ML models forced by different numbers of observation stations were tested, showing that a single wind station alone as an input feature produced a reasonably accurate wave height model. However, the wave height model accuracy increased as more wind input data was included from around the lake, largely plateauing beyond the inclusion of four stations that spanned Lake Michigan's southern basin. The optimized model lookback period was found to be 10 hours for all models, suggestive of a fetch-limited temporal coupling between the wind observations and nearshore waves. The ML framework offers a promising avenue for utilizing historical wind records worldwide to extend wave height time series for nearshore locations, particularly in enclosed and semi-enclosed basins where waves are strongly linked to local winds.
△ Less
Submitted 29 December, 2023; v1 submitted 25 September, 2023;
originally announced September 2023.
