Search for: All records

Abstract In recent decades, there has been a significant increase in annual area burned in California’s Sierra Nevada mountains. This rise in fire activity has prompted the need to understand how historical forest management practices affect fuel composition and emissions. Here we examined the total carbon (TC) concentration and radiocarbon abundance (Δ 14 C) of particulate matter (PM) emitted by the KNP Complex Fire, which occurred during California’s 2021 wildfire season and affected several groves of giant sequoia trees in the southern Sierra Nevada. During a 26 h sampling period, we measured concentrations of fine airborne PM (PM 2.5 ), as well as dry air mole fractions of carbon monoxide (CO) and methane (CH 4 ), using a ground-based mobile laboratory. We also collected filter samples of PM 2.5 for analysis of TC concentration and Δ 14 C. High correlation among PM 2.5 , CO, and CH 4 time series confirmed that our PM 2.5 measurements captured variability in wildfire emissions. Using a Keeling plot approach, we determined that the mean Δ 14 C of PM 2.5 was 111.6 ± 7.7‰ ( n = 12), which was considerably enriched relative to atmospheric carbon dioxide in the northern hemisphere in 2021more »(−3.2 ± 1.4‰). Combining these Δ 14 C data with a steady-state one-box ecosystem model, we estimated that the mean age of fuels combusted in the KNP Complex Fire was 40 years, with a range of 29–57 years. These results provide evidence for emissions originating from woody biomass, larger-diameter fine fuels, and coarse woody debris that have accumulated over multiple decades. This is consistent with independent field observations that indicate high fire intensity contributed to widespread giant sequoia mortality. With the expanded use of prescribed fires planned over the next decade in California to mitigate wildfire impacts, our measurement approach has the potential to provide regionally-integrated estimates of the effectiveness of fuel treatment programs.« less

Quantifying the similarity or distance between time series, processes, signals, and trajectories is a task-specific problem and remains a challenge for many applications. The simplest measure, meaning the Euclidean distance, is often dismissed because of its sensitivity to noise and the curse of dimensionality. Therefore, elastic mappings (such as DTW, LCSS, ED) are often utilized instead. However, these measures are not metric functions, and more importantly, they must deal with the challenges intrinsic to point-to-point mappings, such as pathological alignment. In this paper, we adopt an object-similarity measure, namely Multiscale Intersection over Union (MIoU), for measuring the distance/similarity between time series. We call the new measure TS-MIoU. Unlike the most popular time series similarity measures, TS-MIoU does not rely on a point-to-point mapping, and therefore, circumvents all respective challenges. We show that TS-MIoU is indeed a metric function, especially that it holds the triangle inequality axiom, and therefore can take advantage of indexing algorithms without a lower bounding. We further show that its sensitivity to noise is adjustable, which makes it a strong alternative to the Euclidean distance while not suffering from the curse of dimensionality. Our proof-of-concept experiments on over 100 UCR datasets show that TS-MIoU can fill themore »gap between the unforgiving strictness of the ℓp-norm measures, and the mapping challenges of elastic measures.« less

Despite the success of fully-supervised human skeleton sequence modeling, utilizing self-supervised pre-training for skeleton sequence representation learning has been an active field because acquiring task-specific skeleton annotations at large scales is difficult. Recent studies focus on learning video-level temporal and discriminative information using contrastive learning, but overlook the hierarchical spatial-temporal nature of human skeletons. Different from such superficial supervision at the video level, we propose a self-supervised hierarchical pre-training scheme incorporated into a hierarchical Transformer-based skeleton sequence encoder (Hi-TRS), to explicitly capture spatial, short-term, and long-term temporal dependencies at frame, clip, and video levels, respectively. To evaluate the proposed self-supervised pre-training scheme with Hi-TRS, we conduct extensive experiments covering three skeleton-based downstream tasks including action recognition, action detection, and motion prediction. Under both supervised and semi-supervised evaluation protocols, our method achieves the state-of-the-art performance. Additionally, we demonstrate that the prior knowledge learned by our model in the pre-training stage has strong transfer capability for different downstream tasks.