Reconstructed annual glacier surface mass balance in theányêmaqên Mountains,Yellow River source,based on snow line altitude

Annual mass balance is an important factor that reflects glacier change and glacier meltwater resources.In this study,we analyzed the changes in glacier area,snow line altitude(SLA)and surface elevation in theányêmaqên Mountain region using multisource remote sensing data.Then,the annual mass balance of two glaciers was reconstructed by using SLA-mass-balance gradient method.The results showed that the glacier area in theányêmaqên Mountains decreased by 29.4 km2from 1985 to 2017.The average SLAs of the Halong Glacier and Yehelong Glacier were approximately 5290 m and 5188 m,respectively.The glacier mass balance for the two glaciers from 1990 to 2020 was-0.71 m w.e.a^(-1) and-0.63 m w.e.a^(-1),respectively.Our results indicate that SLA is an important indicator of glacier changes,and a long sequence of SLAs can more accurately reconstruct the glacier mass balance of the glacier.The mean annual glacial meltwater-fed streamflow is 1.45×10^(7)m^(3) and 1.12×10^(7)m^(3),respectively.Sensitivity analysis indicates that summer air temperature plays a leading role in regard to the influential climatic factors of glacial retreat in theányêmaqên Mountains.This highlights the potential of the methodology for application on reconstructing annual glacier surface mass balance at larger scales without direct measurements.

Characterization of contemporary aeolian dust deposition on mountain glaciers of western China

From 2008 to 2010, a total of 15 snow pit samples were collected from 13 mountain glaciers in western China. In this study these samples are used to determine the spatial distribution of insoluble particle concentrations and dust deposition fluxes in western China. The results show that the mass concentrations of insoluble particles exhibit high spatial variation and strongly decrease （by a factor of approximately 50） from the north （Tienshan Mountains） to the south （Himalayas）. However, the insoluble particles concentrations at the southeastern Tibetan Plateau （TP） sites are also high and ap- proximately 30 times greater than those in the Himalayas. The spatial distribution of the dust flux is similar to that of the mass concentrations; however, the high dust deposition rate in the southeastern TP is very significant as a result of the extensive snow accumulation （precipitation） in this region. The average sizes of the insoluble particles at each site generally exhibit bimodal distributions with peaks at approximately 5 μm and 10 μm, which can be explained as re- sulting from dust emissions from regional and local sources, respectively. The enrichment factors for most of the elements measured in insoluble particles are less than 10 at all of the study sites, indicating primarily crustal sources. However, the sites located in the peripheral mountains of western China, such as the Tienshan Mountains and the Himalayas, are characterized by high levels of certain enrichment elements （e.g., Cu, Zn, Cr, and V） indicative of sources related to the long-range transport of pollutants.

Glacier Changes and Their Effects on Northwest China

Though being considered strategically important in matters of national defense based on its abundance of natural resources,Northwest China is one of the most ecologically vulnerable areas in the country. As one of the region's important water sources,glaciers have gone through major changes due to climate change. An analysis of research results over the past 60 years reveals that the glaciers have been retreating in general to some degree in large river basins in Northwest China,but the rate of retreating varied largely among different regions,ranging from 0. 01% ·a^(-1) to 0. 75% ·a^(-1). Specifically,glaciers around the Lantsang River Basin and the Altay Mountains have been retreating fastest. The retreat of glaciers has already caused an increase in runoff in mountainous areas of Northwest China. Meanwhile,increases in the frequency of glacial flash floods and in the flood discharge will further enhance the probability of glacial lake outburst floods and ensuing disasters.

A user-based taxonomy for deep learning visualization

Deep learning has achieved impressive success in a variety of tasks and is developing rapidly in recent years.The problem of understanding the deep learning models has become an issue for the development of deep learning,for example,in domains like medicine and finance which require interpretable models.While it is challenging to analyze and interpret complicated deep neural networks,visualization is good at bridging between abstract data and intuitive representations.Visual analytics for deep learning is a rapidly growing research field.To help users better understand this field,we present a mini-survey including a user-based taxonomy that covers state-of-the-art works of the field.Regarding the requirements of different types of users(beginners,practitioners,developers,and experts),we categorize the methods and tools by four visualization goals respectively focusing on teaching deep learning concepts,architecture assessment,tools for debugging and improving models,and visual explanation.Notably,we present a table consisting of the name of the method or tool,the year,the visualization goal,and the types of networks to which the method or tool can be applied,to assist users in finding available tools and methods quickly.To emphasize the importance of visual explanation for deep learning,we introduce the studies in this research field in detail.

Dynamic ocean inverse modeling based on differentiable rendering

Learning and inferring underlying motion patterns of captured 2D scenes and then re-creating dynamic evolution consistent with the real-world natural phenomena have high appeal for graphics and animation.To bridge the technical gap between virtual and real environments,we focus on the inverse modeling and reconstruction of visually consistent and property-verifiable oceans,taking advantage of deep learning and differentiable physics to learn geometry and constitute waves in a self-supervised manner.First,we infer hierarchical geometry using two networks,which are optimized via the differentiable renderer.We extract wave components from the sequence of inferred geometry through a network equipped with a differentiable ocean model.Then,ocean dynamics can be evolved using the reconstructed wave components.Through extensive experiments,we verify that our new method yields satisfactory results for both geometry reconstruction and wave estimation.Moreover,the new framework has the inverse modeling potential to facilitate a host of graphics applications,such as the rapid production of physically accurate scene animation and editing guided by real ocean scenes.