(Online First)Applying high-resolution visible imagery to satellite melt pond fraction retrieval: A neural network approach

Abstract

During summer, melt ponds have a significant influence on Arctic sea-ice albedo. The melt pond fraction (MPF) also has the ability to forecast the Arctic sea-ice in a certain period. It is important to retrieve accurate melt pond fraction (MPF) from satellite data for Arctic research. This paper proposes a satellite MPF retrieval model based on the multi-layer neural network, named MPF-NN. Our model uses multi-spectral satellite data as model input and MPF information from multi-site and multiperiod visible imagery as prior knowledge for modeling. It can effectively model melt ponds evolution of different regions and periods over the Arctic. Evaluation results show that the MPF retrieved from MODIS data using the proposed model has an RMSE of 3.91% and a correlation coefficient of 0.73. The seasonal distribution of MPF is also consistent with previous results.Index Terms-Multi-layer neural network, high-resolution imagery, melt pond fraction

Review Article: Earth observations of Melt Ponds on Sea Ice

“…method developed by . These findings are corroborated by Liu et al (2017) and Wright and Polashenski (2020), who concluded that machine learning could improve melt pond retrievals from MODIS over current spectral unmixing techniques. After the most traditional techniques for melt ponds studies, such as thresholding and spectral unmixing, ML started to be used for melt pond classifications, for both optical (e.g.…”

“…There has been a growing number of studies applying machine learning-based approaches to MPF retrievals (e.g. Wright, 2020;Liu et al, 2017;Ding et al, 2020a;Feng et al, 2021;Peng et al, 2022). Wright (2020) proposed that the accuracy of the machine learning methods is better than that of the linear spectral unmixing due to the error caused by fixed reflectance feature (e.g.…”

“…ANNs have been used to retrieve directly MPF values from optical data (e.g. Liu et al, 2017;Ding et al, 2019;Ding et al, 2020a;Feng et al, 2021;Peng et al, 2022) and proving to hold a great promise to address the challenge of retrieving MPF. the image that they want to generate or predict) -spectral unmixing MODIS of three classes.…”

Investigation of spatiotemporal variability of melt pond fraction and its relationship with sea ice extent during 2000–2017 using a new data

“…Recently, ANN has been used to retrieve sea ice parameters (Rösel et al, 2012;Liu et al, 2017). It is shown that ANN has the potential to learn the complex relationship between sea ice parameters and input data.…”

“…In many regions, the difference of the averaged MPF between the MODIS retrieval and AMSR-E retrieval during 2002-2011 is less than 5% (Tanaka et al, 2016). More recently, Liu et al (2017) tested an alternative way for the MPF retrieval in summer 2008 by using artificial neural network (ANN). The visible imagery from the National Snow and Ice Data Center (NSIDC) are used as prior knowledge for the network training.…”