2024 Vol. 43, No. 9
Display Method:
2024, 43(9): 1-10.
doi: 10.1007/s13131-024-2352-5
Abstract:
The tide plays a pivotal role in the ocean, affecting the global ocean circulation and supplying the bulk of the energy for the global meridional overturning circulation. To further investigate internal tides and their impacts on circulation, it is imperative to incorporate tidal forcing into the eddy-resolving global ocean circulation model. In this study, we successfully incorporated explicit tides (eight major constituents) into a global eddy-resolving general ocean circulation model and evaluated its tidal simulation ability. We obtained harmonic constants by analyzing sea surface height through tidal harmonic analysis and compared them with the analysis data Topex Poseidon Cross-Overs v9 (TPXO9), the open ocean tide dataset from 102 open-ocean tide observations, and tide gauge stations from World Ocean Circulation Experiment. The results demonstrated that the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics/Institute of Atmospheric Physics (LASG/IAP) Climate System Ocean Model 3.0 (LICOM3.0) effectively simulated tides, with errors predominantly occurring in nearshore regions. The tidal amplitude simulated in LICOM3.0 was greater than that of TPXO9, and these high-amplitude areas exhibited greater errors. The amplitude error of the M2 constituent was larger, while the phase error of the K1 constituent was more significant. Furthermore, we further compared our results with those from other models.
The tide plays a pivotal role in the ocean, affecting the global ocean circulation and supplying the bulk of the energy for the global meridional overturning circulation. To further investigate internal tides and their impacts on circulation, it is imperative to incorporate tidal forcing into the eddy-resolving global ocean circulation model. In this study, we successfully incorporated explicit tides (eight major constituents) into a global eddy-resolving general ocean circulation model and evaluated its tidal simulation ability. We obtained harmonic constants by analyzing sea surface height through tidal harmonic analysis and compared them with the analysis data Topex Poseidon Cross-Overs v9 (TPXO9), the open ocean tide dataset from 102 open-ocean tide observations, and tide gauge stations from World Ocean Circulation Experiment. The results demonstrated that the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics/Institute of Atmospheric Physics (LASG/IAP) Climate System Ocean Model 3.0 (LICOM3.0) effectively simulated tides, with errors predominantly occurring in nearshore regions. The tidal amplitude simulated in LICOM3.0 was greater than that of TPXO9, and these high-amplitude areas exhibited greater errors. The amplitude error of the M2 constituent was larger, while the phase error of the K1 constituent was more significant. Furthermore, we further compared our results with those from other models.
2024, 43(9): 11-25.
doi: 10.1007/s13131-024-2358-z
Abstract:
The three-dimensional displacements caused by ocean loading effects are significant enough to impact spatial geodetic measurements on sub-daily or longer timescales, particularly in the vertical direction. Currently, most tide models incorporate the distribution of vertical displacement loading tides; however, their accuracy has not been assessed for the equatorial and Indian Ocean regions. Global Positioning System (GPS) observations provide high-precision data on sea-level changes, enabling the assessment of the accuracy and reliability of vertical displacement tide models. However, because the tidal period of the K2 constituent is almost identical to the orbital period of GPS constellations, the estimation of the K2 tidal constituent from GPS observations is not satisfactory. In this study, the principle of smoothness is employed to correct the systematic error in K2 estimates in GPS observations through quadratic fitting. Using the adjusted harmonic constants from 31 GPS stations for the equatorial and Indian Ocean, the accuracy of eight major constituents from five global vertical displacement tide models (FES2014, EOT11a, GOT4.10c, GOT4.8, and NAO.99b) is evaluated for the equatorial and Indian Ocean. The results indicate that the EOT11a and FES2014 models exhibit higher accuracy in the vertical displacement tide models for the equatorial and Indian Ocean, with root sum squares errors of 2.29 mm and 2.34 mm, respectively. Furthermore, a brief analysis of the vertical displacement tide distribution characteristics of the eight major constituents for the equatorial and Indian Ocean was conducted using the EOT11a model.
The three-dimensional displacements caused by ocean loading effects are significant enough to impact spatial geodetic measurements on sub-daily or longer timescales, particularly in the vertical direction. Currently, most tide models incorporate the distribution of vertical displacement loading tides; however, their accuracy has not been assessed for the equatorial and Indian Ocean regions. Global Positioning System (GPS) observations provide high-precision data on sea-level changes, enabling the assessment of the accuracy and reliability of vertical displacement tide models. However, because the tidal period of the K2 constituent is almost identical to the orbital period of GPS constellations, the estimation of the K2 tidal constituent from GPS observations is not satisfactory. In this study, the principle of smoothness is employed to correct the systematic error in K2 estimates in GPS observations through quadratic fitting. Using the adjusted harmonic constants from 31 GPS stations for the equatorial and Indian Ocean, the accuracy of eight major constituents from five global vertical displacement tide models (FES2014, EOT11a, GOT4.10c, GOT4.8, and NAO.99b) is evaluated for the equatorial and Indian Ocean. The results indicate that the EOT11a and FES2014 models exhibit higher accuracy in the vertical displacement tide models for the equatorial and Indian Ocean, with root sum squares errors of 2.29 mm and 2.34 mm, respectively. Furthermore, a brief analysis of the vertical displacement tide distribution characteristics of the eight major constituents for the equatorial and Indian Ocean was conducted using the EOT11a model.
2024, 43(9): 26-34.
doi: 10.1007/s13131-024-2357-0
Abstract:
The frontogenetic processes of a submesoscale cold filament driven by the thermal convection turbulence are studied by a non-hydrostatic large eddy simulation. The results show that the periodic changes in the direction of the cross-filament secondary circulations are induced by the inertial oscillation. The change in the direction of the secondary circulations induces the enhancement and reduction of the horizontal temperature gradient during the former and later inertial period, which indicates that the frontogenetical processes of the cold filament include both of frontogenesis and frontolysis. The structure of the cold filament may be broken and restored by frontogenesis and frontolysis, respectively. The magnitude of the down-filament currents has a periodic variation, while its direction is unchanged with time. The coupling effect of the turbulent mixing and the frontogenesis and frontolysis gradually weakens the temperature gradient of the cold filament with time, which reduces frontogenetical intensity and enlarges the width of cold filament.
The frontogenetic processes of a submesoscale cold filament driven by the thermal convection turbulence are studied by a non-hydrostatic large eddy simulation. The results show that the periodic changes in the direction of the cross-filament secondary circulations are induced by the inertial oscillation. The change in the direction of the secondary circulations induces the enhancement and reduction of the horizontal temperature gradient during the former and later inertial period, which indicates that the frontogenetical processes of the cold filament include both of frontogenesis and frontolysis. The structure of the cold filament may be broken and restored by frontogenesis and frontolysis, respectively. The magnitude of the down-filament currents has a periodic variation, while its direction is unchanged with time. The coupling effect of the turbulent mixing and the frontogenesis and frontolysis gradually weakens the temperature gradient of the cold filament with time, which reduces frontogenetical intensity and enlarges the width of cold filament.
2024, 43(9): 35-53.
doi: 10.1007/s13131-024-2360-5
Abstract:
It is found that the winter (December–February) barrier layer (BL) in the Bay of Bengal (BoB) acts as a dynamical thermostat, modulating the subsequent summer BoB sea surface temperature (SST) variability and potentially affecting the Indian summer monsoon (ISM) onset and associated rainfall variability. In the years when the prior winter BL is anomalously thick, anomalous sea surface cooling caused by intensified latent heat flux loss appears in the BoB starting in October and persists into the following year by positive cloud-SST feedback. During January–March, the vertical entrainment of warmer subsurface water induced by the anomalously thick BL acts to damp excessive cooling of the sea surface caused by atmospheric forcing and favors the development of deep atmospheric convection over the BoB. During March–May, the thinner mixed layer linked to the anomalously thick BL allows more shortwave radiation to penetrate below the mixed layer. This tends to maintain existing cold SST anomalies, advancing the onset of ISM and enhancing June ISM precipitation through an increase in the land-sea tropospheric thermal contrast. We also find that most of the coupled model intercomparison project phase 5 (CMIP5) models fail to reproduce the observed relationship between June ISM rainfall and the prior winter BL thickness. This may be attributable to their difficulties in realistically simulating the winter BL in the BoB and ISM precipitation. The present results indicate that it is important to realistically capture the winter BL of the BoB in climate models for improving the simulation and prediction of ISM.
It is found that the winter (December–February) barrier layer (BL) in the Bay of Bengal (BoB) acts as a dynamical thermostat, modulating the subsequent summer BoB sea surface temperature (SST) variability and potentially affecting the Indian summer monsoon (ISM) onset and associated rainfall variability. In the years when the prior winter BL is anomalously thick, anomalous sea surface cooling caused by intensified latent heat flux loss appears in the BoB starting in October and persists into the following year by positive cloud-SST feedback. During January–March, the vertical entrainment of warmer subsurface water induced by the anomalously thick BL acts to damp excessive cooling of the sea surface caused by atmospheric forcing and favors the development of deep atmospheric convection over the BoB. During March–May, the thinner mixed layer linked to the anomalously thick BL allows more shortwave radiation to penetrate below the mixed layer. This tends to maintain existing cold SST anomalies, advancing the onset of ISM and enhancing June ISM precipitation through an increase in the land-sea tropospheric thermal contrast. We also find that most of the coupled model intercomparison project phase 5 (CMIP5) models fail to reproduce the observed relationship between June ISM rainfall and the prior winter BL thickness. This may be attributable to their difficulties in realistically simulating the winter BL in the BoB and ISM precipitation. The present results indicate that it is important to realistically capture the winter BL of the BoB in climate models for improving the simulation and prediction of ISM.
2024, 43(9): 54-69.
doi: 10.1007/s13131-024-2361-4
Abstract:
Shipboard radiosonde soundings are important for detecting and quantifying the multiscale variability of atmosphere-ocean interactions associated with mass exchanges. This study evaluated the accuracies of shipboard Global Positioning System (GPS) soundings in the eastern tropical Indian Ocean and South China Sea through a simultaneous balloon-borne inter-comparison of different radiosonde types. Our results indicate that the temperature and relative humidity (RH) measurements of GPS-TanKong (GPS-TK) radiosonde (used at most stations before 2012) have larger biases than those of ChangFeng-06-A (CF-06-A) radiosonde (widely used in current observation) when compared to reference data from Vaisala RS92-SGP radiosonde, with a warm bias of 5℃ and dry bias of 10% during daytimes, and a cooling bias of –0.8℃ and a moist bias of 6% during nighttime. These systematic biases are primarily attributed to the radiation effects and altitude deviation. An empirical correction algorithm was developed to retrieve the atmospheric temperature and RH profiles. The corrected profiles agree well with that of RS92-SGP, except for uncertainties of CF-06-A in the stratosphere. These correction algorithms were applied to the GPS-TK historical sounding records, reducing biases in the corrected temperature and RH profiles when compared to radio occultation data. The correction of GPS-TK historical records illustrated an improvement in capturing the marine atmospheric structure, with more accurate atmospheric boundary layer height, convective available potential energy, and convective inhibition in the tropical ocean. This study contributes significantly to improving the quality of GPS radiosonde soundings and promotes the sharing of observation in the eastern tropical Indian Ocean and South China Sea.
Shipboard radiosonde soundings are important for detecting and quantifying the multiscale variability of atmosphere-ocean interactions associated with mass exchanges. This study evaluated the accuracies of shipboard Global Positioning System (GPS) soundings in the eastern tropical Indian Ocean and South China Sea through a simultaneous balloon-borne inter-comparison of different radiosonde types. Our results indicate that the temperature and relative humidity (RH) measurements of GPS-TanKong (GPS-TK) radiosonde (used at most stations before 2012) have larger biases than those of ChangFeng-06-A (CF-06-A) radiosonde (widely used in current observation) when compared to reference data from Vaisala RS92-SGP radiosonde, with a warm bias of 5℃ and dry bias of 10% during daytimes, and a cooling bias of –0.8℃ and a moist bias of 6% during nighttime. These systematic biases are primarily attributed to the radiation effects and altitude deviation. An empirical correction algorithm was developed to retrieve the atmospheric temperature and RH profiles. The corrected profiles agree well with that of RS92-SGP, except for uncertainties of CF-06-A in the stratosphere. These correction algorithms were applied to the GPS-TK historical sounding records, reducing biases in the corrected temperature and RH profiles when compared to radio occultation data. The correction of GPS-TK historical records illustrated an improvement in capturing the marine atmospheric structure, with more accurate atmospheric boundary layer height, convective available potential energy, and convective inhibition in the tropical ocean. This study contributes significantly to improving the quality of GPS radiosonde soundings and promotes the sharing of observation in the eastern tropical Indian Ocean and South China Sea.
2024, 43(9): 70-80.
doi: 10.1007/s13131-024-2350-7
Abstract:
Typhoons in the western Pacific have a significant impact on the transport of heat, salt and particles through the Luzon Strait. However, there are very limited field observations of this impact because of extreme difficulties and even dangers for ship-based measurements during the rough weather. Here, we present the preliminary results from analyzing a dataset collected by a glider deployed west of the Luzon Strait a few days prior to the arrival of typhoon MITAG. The gilder data revealed an abnormally salinity (>34.8) subsurface water apparently sourced from Kuroshio intrusion during the typhoon. When typhoon MITAG traveled on the east of the Luzon Strait, the positive wind stress curl strengthened the cyclonic eddy and weakened the anti-cyclonic eddy. This led to a slowdown of Kuroshio and made its intrusion easier. The main axis of the Kuroshio at the northern part of the strait shifted westward after the typhoon and did not return to its original position until a week later. The Ekman transport from persistent northerly wind of typhoon MITAG was significant, but its importance in enhancing the Kuroshio intrusion is only secondary relative to the eddies variations.
Typhoons in the western Pacific have a significant impact on the transport of heat, salt and particles through the Luzon Strait. However, there are very limited field observations of this impact because of extreme difficulties and even dangers for ship-based measurements during the rough weather. Here, we present the preliminary results from analyzing a dataset collected by a glider deployed west of the Luzon Strait a few days prior to the arrival of typhoon MITAG. The gilder data revealed an abnormally salinity (>34.8) subsurface water apparently sourced from Kuroshio intrusion during the typhoon. When typhoon MITAG traveled on the east of the Luzon Strait, the positive wind stress curl strengthened the cyclonic eddy and weakened the anti-cyclonic eddy. This led to a slowdown of Kuroshio and made its intrusion easier. The main axis of the Kuroshio at the northern part of the strait shifted westward after the typhoon and did not return to its original position until a week later. The Ekman transport from persistent northerly wind of typhoon MITAG was significant, but its importance in enhancing the Kuroshio intrusion is only secondary relative to the eddies variations.
2024, 43(9): 81-92.
doi: 10.1007/s13131-024-2364-1
Abstract:
Sediment-laden sea ice plays an important role in Arctic sediment transport and biogeochemical cycles, as well as the shortwave radiation budget and melt onset of ice surface. However, at present, there is a lack of efficient observation approach from both space and in situ for the coverage of Arctic sediment-laden sea ice. Thus, both spatial distribution and long-term changes in area fraction of such ice floes are still unclear. This study proposes a new classification method to extract Arctic sediment-laden sea ice on the basic of the difference in spectral characteristics between sediment-laden sea ice and clean sea ice in the visible band using the MOD09A1 data with the resolution of 500 m, and obtains its area fraction over the pan Arctic Ocean during 2000−2021. Compared with Landsat-8 true color verification images with a resolution of 30 m, the overall accuracy of our classification method is 92.3%, and the Kappa coefficient is 0.84. The impact of clouds on the results of recognition and spatiotemporal changes of sediment-laden sea ice is relatively small from June to July, compared to that in May or August. Spatially, sediment-laden sea ice mostly appears over the marginal seas of the Arctic Ocean, especially the continental shelf of Chukchi Sea and the Siberian seas. Associated with the retreat of Arctic sea ice extent, the total area of sediment-laden sea ice in June–July also shows a significant decreasing trend of 8.99 × 104 km2 per year. The occurrence of sediment-laden sea ice over the Arctic Ocean in June–July leads to the reduce of surface albedo over the ice-covered ocean by 14.1%. This study will help thoroughly understanding of the role of sediment-laden sea ice in the evolution of Arctic climate system and marine ecological environment, as well as the heat budget and mass balance of sea ice itself.
Sediment-laden sea ice plays an important role in Arctic sediment transport and biogeochemical cycles, as well as the shortwave radiation budget and melt onset of ice surface. However, at present, there is a lack of efficient observation approach from both space and in situ for the coverage of Arctic sediment-laden sea ice. Thus, both spatial distribution and long-term changes in area fraction of such ice floes are still unclear. This study proposes a new classification method to extract Arctic sediment-laden sea ice on the basic of the difference in spectral characteristics between sediment-laden sea ice and clean sea ice in the visible band using the MOD09A1 data with the resolution of 500 m, and obtains its area fraction over the pan Arctic Ocean during 2000−2021. Compared with Landsat-8 true color verification images with a resolution of 30 m, the overall accuracy of our classification method is 92.3%, and the Kappa coefficient is 0.84. The impact of clouds on the results of recognition and spatiotemporal changes of sediment-laden sea ice is relatively small from June to July, compared to that in May or August. Spatially, sediment-laden sea ice mostly appears over the marginal seas of the Arctic Ocean, especially the continental shelf of Chukchi Sea and the Siberian seas. Associated with the retreat of Arctic sea ice extent, the total area of sediment-laden sea ice in June–July also shows a significant decreasing trend of 8.99 × 104 km2 per year. The occurrence of sediment-laden sea ice over the Arctic Ocean in June–July leads to the reduce of surface albedo over the ice-covered ocean by 14.1%. This study will help thoroughly understanding of the role of sediment-laden sea ice in the evolution of Arctic climate system and marine ecological environment, as well as the heat budget and mass balance of sea ice itself.
2024, 43(9): 1-2.
Abstract:
2024, 43(9): 93-104.
doi: 10.1007/s13131-024-2362-3
Abstract:
In the last 10 years (2012–2021), five hypoxic events have been observed in summer in the central Bohai Sea (CBS). Frequent and persistent hypoxia will have an impact on the ecosystem of the CBS. In this paper, historical sea temperature (ST), salinity (SAL), density (Den), and dissolved oxygen (DO) concentration data from three stations in the CBS are analyzed via the linear regression method, and the correlations between the stratification factors (ST, SAL, and Den) and DO concentration are determined. The thresholds of the stratification factors at the three stations in June in the year in which hypoxia occurred were determined and applied to survey data from 29 stations in late May to early June in 2022 in the CBS; this assessment found that the data from 19 stations indicated that hypoxia was about to occur. In August, the survey data showed that 14 out of the 29 stations indicated hypoxic conditions, of which 12 were from the predicted 19 stations, meaning that the estimation accuracy reached 63%. The same approach was applied to data from June 2023. The data for August from a bottom-type online monitoring system in the CBS verified the occurrence of hypoxic events around Sta. M2. The results show that the strength of the seawater stratification plays a leading role in hypoxic events in the summer in the CBS, and the thresholds of the stratification factors can be used to predict the occurrence of hypoxic events.
In the last 10 years (2012–2021), five hypoxic events have been observed in summer in the central Bohai Sea (CBS). Frequent and persistent hypoxia will have an impact on the ecosystem of the CBS. In this paper, historical sea temperature (ST), salinity (SAL), density (Den), and dissolved oxygen (DO) concentration data from three stations in the CBS are analyzed via the linear regression method, and the correlations between the stratification factors (ST, SAL, and Den) and DO concentration are determined. The thresholds of the stratification factors at the three stations in June in the year in which hypoxia occurred were determined and applied to survey data from 29 stations in late May to early June in 2022 in the CBS; this assessment found that the data from 19 stations indicated that hypoxia was about to occur. In August, the survey data showed that 14 out of the 29 stations indicated hypoxic conditions, of which 12 were from the predicted 19 stations, meaning that the estimation accuracy reached 63%. The same approach was applied to data from June 2023. The data for August from a bottom-type online monitoring system in the CBS verified the occurrence of hypoxic events around Sta. M2. The results show that the strength of the seawater stratification plays a leading role in hypoxic events in the summer in the CBS, and the thresholds of the stratification factors can be used to predict the occurrence of hypoxic events.
2024, 43(9): 105-109.
doi: 10.1007/s13131-024-2355-2
Abstract:
During August and September 2023, three giant icebergs, each bigger than Paris, successively grazed Clarence Island in the northeast of the Antarctic Peninsula, a home to a population of over 100 000 penguins. This incident may serve as a clarion call for the increasing iceberg calving due to global warming and its subsequent impact on the Antarctic ecosystem. Here we investigate this unexpected event and employ historical records and probabilistic analyses of iceberg grounding to assess the degree of impact on penguin colonies of Clarence Island. Among the eleven colonies, there is one with low impact, eight with medium impact, and two with high impact. The low-impact colony, Cape Lloyd, is located in the northern part of the island, while the high-impact colonies, False Ridge and Pink Pool, are in the southeast. The eight medium-impact colonies are distributed along both the eastern and western coasts of the island. This study provides essential support for evaluating the impact of iceberg activity on penguin colonies. We argue that penguin colonies located in areas prone to iceberg drift, such as Clarence Island, may become more vulnerable to the heightened risk of iceberg collisions or groundings in the warming future. Therefore, we hope the public will become more aware of the grave impacts of climate change on penguins and underscore the urgent need for effective conservation strategies.
During August and September 2023, three giant icebergs, each bigger than Paris, successively grazed Clarence Island in the northeast of the Antarctic Peninsula, a home to a population of over 100 000 penguins. This incident may serve as a clarion call for the increasing iceberg calving due to global warming and its subsequent impact on the Antarctic ecosystem. Here we investigate this unexpected event and employ historical records and probabilistic analyses of iceberg grounding to assess the degree of impact on penguin colonies of Clarence Island. Among the eleven colonies, there is one with low impact, eight with medium impact, and two with high impact. The low-impact colony, Cape Lloyd, is located in the northern part of the island, while the high-impact colonies, False Ridge and Pink Pool, are in the southeast. The eight medium-impact colonies are distributed along both the eastern and western coasts of the island. This study provides essential support for evaluating the impact of iceberg activity on penguin colonies. We argue that penguin colonies located in areas prone to iceberg drift, such as Clarence Island, may become more vulnerable to the heightened risk of iceberg collisions or groundings in the warming future. Therefore, we hope the public will become more aware of the grave impacts of climate change on penguins and underscore the urgent need for effective conservation strategies.
2024, 43(9): 110-121.
doi: 10.1007/s13131-024-2356-1
Abstract:
Mangroves are indispensable to coastlines, maintaining biodiversity, and mitigating climate change. Therefore, improving the accuracy of mangrove information identification is crucial for their ecological protection. Aiming at the limited morphological information of synthetic aperture radar (SAR) images, which is greatly interfered by noise, and the susceptibility of optical images to weather and lighting conditions, this paper proposes a pixel-level weighted fusion method for SAR and optical images. Image fusion enhanced the target features and made mangrove monitoring more comprehensive and accurate. To address the problem of high similarity between mangrove forests and other forests, this paper is based on the U-Net convolutional neural network, and an attention mechanism is added in the feature extraction stage to make the model pay more attention to the mangrove vegetation area in the image. In order to accelerate the convergence and normalize the input, batch normalization (BN) layer and Dropout layer are added after each convolutional layer. Since mangroves are a minority class in the image, an improved cross-entropy loss function is introduced in this paper to improve the model’s ability to recognize mangroves. The AttU-Net model for mangrove recognition in high similarity environments is thus constructed based on the fused images. Through comparison experiments, the overall accuracy of the improved U-Net model trained from the fused images to recognize the predicted regions is significantly improved. Based on the fused images, the recognition results of the AttU-Net model proposed in this paper are compared with its benchmark model, U-Net, and the Dense-Net, Res-Net, and Seg-Net methods. The AttU-Net model captured mangroves’ complex structures and textural features in images more effectively. The average OA, F1-score, and Kappa coefficient in the four tested regions were 94.406%, 90.006%, and 84.045%, which were significantly higher than several other methods. This method can provide some technical support for the monitoring and protection of mangrove ecosystems.
Mangroves are indispensable to coastlines, maintaining biodiversity, and mitigating climate change. Therefore, improving the accuracy of mangrove information identification is crucial for their ecological protection. Aiming at the limited morphological information of synthetic aperture radar (SAR) images, which is greatly interfered by noise, and the susceptibility of optical images to weather and lighting conditions, this paper proposes a pixel-level weighted fusion method for SAR and optical images. Image fusion enhanced the target features and made mangrove monitoring more comprehensive and accurate. To address the problem of high similarity between mangrove forests and other forests, this paper is based on the U-Net convolutional neural network, and an attention mechanism is added in the feature extraction stage to make the model pay more attention to the mangrove vegetation area in the image. In order to accelerate the convergence and normalize the input, batch normalization (BN) layer and Dropout layer are added after each convolutional layer. Since mangroves are a minority class in the image, an improved cross-entropy loss function is introduced in this paper to improve the model’s ability to recognize mangroves. The AttU-Net model for mangrove recognition in high similarity environments is thus constructed based on the fused images. Through comparison experiments, the overall accuracy of the improved U-Net model trained from the fused images to recognize the predicted regions is significantly improved. Based on the fused images, the recognition results of the AttU-Net model proposed in this paper are compared with its benchmark model, U-Net, and the Dense-Net, Res-Net, and Seg-Net methods. The AttU-Net model captured mangroves’ complex structures and textural features in images more effectively. The average OA, F1-score, and Kappa coefficient in the four tested regions were 94.406%, 90.006%, and 84.045%, which were significantly higher than several other methods. This method can provide some technical support for the monitoring and protection of mangrove ecosystems.
2024, 43(9): 122-135.
doi: 10.1007/s13131-024-2351-6
Abstract:
The long-term dynamic evolution and underlying mechanisms of coastal landscape pattern stability, driven by strong anthropogenic interference and consequently climate change, are topics of major interest in national and international scientific research. Guangdong Province, located in southeastern China, has been undergoing rapid urbanization over several decades. In this study, we quantitatively determined the scale threshold characteristics of coastal landscape pattern stability in Guangdong Province, from the dual perspective of spatial heterogeneity and spatial autocorrelation. An analysis of the spatiotemporal evolution of the coastal landscape was conducted after the optical scale was determined. Then, we applied the geodetector statistical method to quantitatively explore the mechanisms underlying coastal landscape pattern stability. Based on the inflection point of landscape metrics and the maximum value of the Moran Ⅰ index, the optimal scale for analyzing coastal landscape pattern stability in Guangdong Province was 240 m × 240 m. Within the past several decades, coastal landscape pattern stability increased slightly and then decreased, with a turning point around 2005. The most significant variations in coastal landscape pattern stability were observed in the transition zone of rural-urban expansion. A q-statistics analysis showed that the explanatory power of paired factors was greater than that of a single driving factor; the paired factors with the greatest impact on coastal landscape pattern stability in Guangdong Province were the change in gross industrial output and change in average annual precipitation from 2010 to 2015, based on a q value of 0.604. These results will contribute to future efforts to achieve sustainable coastal development and provide a scientific basis and technical support for the rational planning and utilization of resources in large estuarine areas, including marine disaster prevention and seawall ecological restoration.
The long-term dynamic evolution and underlying mechanisms of coastal landscape pattern stability, driven by strong anthropogenic interference and consequently climate change, are topics of major interest in national and international scientific research. Guangdong Province, located in southeastern China, has been undergoing rapid urbanization over several decades. In this study, we quantitatively determined the scale threshold characteristics of coastal landscape pattern stability in Guangdong Province, from the dual perspective of spatial heterogeneity and spatial autocorrelation. An analysis of the spatiotemporal evolution of the coastal landscape was conducted after the optical scale was determined. Then, we applied the geodetector statistical method to quantitatively explore the mechanisms underlying coastal landscape pattern stability. Based on the inflection point of landscape metrics and the maximum value of the Moran Ⅰ index, the optimal scale for analyzing coastal landscape pattern stability in Guangdong Province was 240 m × 240 m. Within the past several decades, coastal landscape pattern stability increased slightly and then decreased, with a turning point around 2005. The most significant variations in coastal landscape pattern stability were observed in the transition zone of rural-urban expansion. A q-statistics analysis showed that the explanatory power of paired factors was greater than that of a single driving factor; the paired factors with the greatest impact on coastal landscape pattern stability in Guangdong Province were the change in gross industrial output and change in average annual precipitation from 2010 to 2015, based on a q value of 0.604. These results will contribute to future efforts to achieve sustainable coastal development and provide a scientific basis and technical support for the rational planning and utilization of resources in large estuarine areas, including marine disaster prevention and seawall ecological restoration.
2024, 43(9): 136-144.
doi: 10.1007/s13131-024-2359-y
Abstract:
Wave information retrieval from videos captured by a single camera has been increasingly applied in marine observation. However, when the camera observes ocean waves at low grazing angles, the accurate extraction of wave information from videos will be affected by the interference of the fine ripples on the sea surface. To solve this problem, this study develops a method for estimating peak wave periods from videos captured at low grazing angles. The method extracts the motion of the sea surface texture from the video and obtains the peak wave period via the spectral analysis. The calculation results captured from real-world videos are compared with those obtained from X-band radar inversion and tracking buoy movement, with maximum deviations of 8% and 14%, respectively. The analysis of the results shows that the peak wave period of the method has good stability. In addition, this paper uses a pinhole camera model to convert the displacement of the texture from pixel height to actual height and performs moving average filtering on the displacement of the texture, thus conducting a preliminary exploration of the inversion of significant wave height. This study helps to extend the application of sea surface videos.
Wave information retrieval from videos captured by a single camera has been increasingly applied in marine observation. However, when the camera observes ocean waves at low grazing angles, the accurate extraction of wave information from videos will be affected by the interference of the fine ripples on the sea surface. To solve this problem, this study develops a method for estimating peak wave periods from videos captured at low grazing angles. The method extracts the motion of the sea surface texture from the video and obtains the peak wave period via the spectral analysis. The calculation results captured from real-world videos are compared with those obtained from X-band radar inversion and tracking buoy movement, with maximum deviations of 8% and 14%, respectively. The analysis of the results shows that the peak wave period of the method has good stability. In addition, this paper uses a pinhole camera model to convert the displacement of the texture from pixel height to actual height and performs moving average filtering on the displacement of the texture, thus conducting a preliminary exploration of the inversion of significant wave height. This study helps to extend the application of sea surface videos.
