Downscaling Snow Depth Mapping by Fusion of Microwave and Optical Remote-Sensing Data Based on Deep Learning

Zhu, Linglong; Zhang, Yonghong; Wang, Jiangeng; Tian, Wei; Liu, Qi; Ma, Guangyi; Kan, Xi; Chu, Ya

doi:10.3390/rs13040584

Cited by 29 publications

(20 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The training process is an iterative process with the aim to minimize a cost function, for example the average squared error between the network outputs and the targets 69 . ANNs have been used for several purposes: among them the estimation of biophysical parameters 67 , 70 , 71 , pattern recognition and downscaling approaches 20 . In this work, a classical feed-forward network with error backpropagation algorithm was adopted.…”

Section: Methodsmentioning

confidence: 99%

“…The use of modeled and/or reanalysis data can have some advantages with respect to ground data and satellite observations, being more consistent in time and space 17 . However, the coarse spatial resolution of the modeled and/or reanalysis data sets resulted in difficulty to well represent the spatial variability of mountain processes 18 – 20 . In this context, downscaling approaches can be used to improve the low resolution modeled data through the use of high resolution data derived from satellite imagery 21 , 22 .…”

Section: Introductionmentioning

confidence: 99%

“…More recent works have started adopting methods based on machine learning, which can easily deal with the non-linearity among variables and large data sets. Among the different machine learning approaches, Artificial Neural Networks (ANNs) have been widely adopted 19 , 20 , 23 .…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Overall negative trends for snow cover extent and duration in global mountain regions over 1982–2020

Notarnicola

2022

Sci Rep

View full text Add to dashboard Cite

Notwithstanding the large availability of data and models, a consistent picture of the snow cover extent and duration changes in global mountain areas is lacking for long-term trends. Here, model data and satellite images are combined by using Artificial Neural Networks to generate a consistent time series from 1982 to 2020 over global mountain areas. The analysis of the harmonized time series over 38 years indicates an overall negative trend of − 3.6% ± 2.7% for yearly snow cover extent and of − 15.1 days ± 11.6 days for snow cover duration. The most affected season by negative trends is winter with an average reduction in snow cover extent of − 11.5% ± 6.9%, and the most affected season by positive changes is spring with an average increase of 10% ± 5.9%, the latter mainly located in High Mountain Asia. The results indicated a shift in the snow regime located between the 80 s and 90 s of the previous century, where the period from 1982 to 1999 is characterized by a higher number of areas with significant changes and a higher rate of changes with respect to the period 2000–2020. This quantification can lead to a more accurate evaluation of the impact on water resources for mountainous communities.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Overall negative trends for snow cover extent and duration in global mountain regions over 1982–2020

Notarnicola

2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…where TP, FN, and FP represent the number of correctly detected, undetected, and incorrectly detected telegraph poles in the results, respectively. The Root Mean Square Error (RMSE) [29] index is utilized to calculate the plane positioning error of the telegraph pole that the algorithm can correctly detect while quantitatively assessing the telegraph pole positioning deviation. The calculation formula is shown in Equation (6).…”

Section: Accuracy Evaluation (1) Telegraph Pole Locatingmentioning

confidence: 99%

Location and Extraction of Telegraph Poles from Image Matching-Based Point Clouds

Wang

et al. 2022

Remote Sensing

View full text Add to dashboard Cite

The monitoring of telegraph poles as essential features supporting overhead distribution network lines is the primary subject of this work. This paper proposes a method for locating and extracting telegraph poles from an image matching-based point cloud. Firstly, the point cloud of the poles is extracted using the planar grid segmentation clustering algorithm and the connected component analysis algorithm of the region grows according to the isolated features of the poles perpendicular to the ground. Secondly, the candidate telegraph poles are located based on the suspension point of the buffer, considering that the top of the pole is connected to the power suspension line. Thirdly, the horizontal projection method of the backbone area is utilized to eliminate the interference of vegetation in the buffer area. Finally, the point cloud of the telegraph pole is extracted through the density-based spatial clustering of applications with noise (DBSCAN) algorithm. The experimental results demonstrate that the average values of Recall, Precision, and F1-score in telegraph pole detection can reach 91.09%, 90.82%, and 90.90%, respectively. The average RMSE value of location deviation is 0.51m. The average value of the F1-score in the telegraph pole extraction is 91.83%, and the average extraction time of a single pole is 0.27s. Accordingly, this method has strong adaptability to areas with lush vegetation and can automatically locate and extract the telegraph pole point cloud with high accuracy, and it can still achieve very high accuracy even under the holes in the data.

show abstract

“…As described in previous relevant papers, machine learning methods could overcome various complex problems existing in large-scale retrievals [27][28][29]. Machine-learning methods can learn and summarize a large number of data and not rely on the understanding of physical processes when modeling [26]. Although the fused snow depth dataset performs well in accuracy assessment via five independent in situ observations, there are still some limitations that warrant further improvements.…”

Section: Limitations Of the Current Studymentioning

confidence: 99%

Snow Depth Fusion Based on Machine Learning Methods for the Northern Hemisphere

Che

Dai

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

In this study, a machine learning algorithm was introduced to fuse gridded snow depth datasets. The input variables of the machine learning method included geolocation (latitude and longitude), topographic data (elevation), gridded snow depth datasets and in situ observations. A total of 29,565 in situ observations were used to train and optimize the machine learning algorithm. A total of five gridded snow depth datasets—Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) snow depth, Global Snow Monitoring for Climate Research (GlobSnow) snow depth, Long time series of daily snow depth over the Northern Hemisphere (NHSD) snow depth, ERA-Interim snow depth and Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2) snow depth—were used as input variables. The first three snow depth datasets are retrieved from passive microwave brightness temperature or assimilation with in situ observations, while the last two are snow depth datasets obtained from meteorological reanalysis data with a land surface model and data assimilation system. Then, three machine learning methods, i.e., Artificial Neural Networks (ANN), Support Vector Regression (SVR), and Random Forest Regression (RFR), were used to produce a fused snow depth dataset from 2002 to 2004. The RFR model performed best and was thus used to produce a new snow depth product from the fusion of the five snow depth datasets and auxiliary data over the Northern Hemisphere from 2002 to 2011. The fused snow-depth product was verified at five well-known snow observation sites. The R2 of Sodankylä, Old Aspen, and Reynolds Mountains East were 0.88, 0.69, and 0.63, respectively. At the Swamp Angel Study Plot and Weissfluhjoch observation sites, which have an average snow depth exceeding 200 cm, the fused snow depth did not perform well. The spatial patterns of the average snow depth were analyzed seasonally, and the average snow depths of autumn, winter, and spring were 5.7, 25.8, and 21.5 cm, respectively. In the future, random forest regression will be used to produce a long time series of a fused snow depth dataset over the Northern Hemisphere or other specific regions.

show abstract

Downscaling Snow Depth Mapping by Fusion of Microwave and Optical Remote-Sensing Data Based on Deep Learning

Cited by 29 publications

References 58 publications

Overall negative trends for snow cover extent and duration in global mountain regions over 1982–2020

Overall negative trends for snow cover extent and duration in global mountain regions over 1982–2020

Location and Extraction of Telegraph Poles from Image Matching-Based Point Clouds

Snow Depth Fusion Based on Machine Learning Methods for the Northern Hemisphere

Contact Info

Product

Resources

About