Multipath is the main systematic error of the Global Navigation Satellite System (GNSS) short baseline positioning. Multipath cannot be eliminated by the double-differenced technique and is difficult to parameterize, which severely restrict the high-precision GNSS positioning application. Based on the spatiotemporal repeatability of multipath, the sidereal filtering in coordinate-domain (SF-CD), the sidereal filtering in observation-domain (SF-OD), and the multipath hemispherical map (MHM) can be used to mitigate the multipath in real-time. However, the multipath model with large matrix for multi-GNSS multipath mitigation is difficult to achieve lightweight calculation and the SF-CD cannot be applied to mitigate the multi-GNSS multipath. In this paper, we propose a new multipath mitigation strategy in the coordinate-domain that shakes off the formation mechanism of multipath, a CNN (convolutional neural network)-LSTM (long short-term memory) method is used to mine the deep multipath features in GNSS coordinate series. Furthermore, multipath will be mitigated in real-time by constantly predicting the value of the next epoch. The experimental results show that the CNN-LSTM effectively mitigates the multi-GNSS multipath. The method can reduce the average RMS (root-mean square) of multi-GNSS positioning errors in the east, north, and vertical directions by 62.3%, 70.8%, and 66.0%. Moreover, comparing with the SF-CD, SF-OD, and MHM, CNN-LSTM can more effectively mitigate the effects of the GPS multipath, and the ability of multipath mitigation is almost not affected over time.

Although there is a consensus that protected areas (PAs) can provide various ecosystem services, it is unclear whether protected areas effectively contribute to the preservation and enhancement of ecosystem services. We conducted a case study of the Hoh Xil Nature Reserve (HXNR) in Qinghai-Tibetan Plateau, China, in order to examine the effectiveness of PA in the conservation of ecosystem services. First, the dynamics of land use/land cover (LULC) were analyzed based on remotely sensed data sets. Then, the ecosystem service value (ESV) in the PA and non-PA were evaluated using a modified benefit transfer method that had been adjusted using normalized difference vegetation index (NDVI). Finally, hotspot analysis was implemented to reveal the ESV changes for the different districts of the PA. The results of the comparison experiment indicate that: (1) The ESV of the HXNR has considerably increased after it was designated as protected, which had been in decline in the previous stage. The ESVs in a near-by non-PA showed opposite results where the values initially increased but then dropped due to urban expansion and desertification. (2) The areas in HXNR with increased ESV significantly outnumbered the areas that had declining values from 1980 to 2018. For the non-PA, the areas that had increased ESV in 1980–1995 saw a decline in value in 1995–2008; moreover, new areas with decreasing ESV emerged in 2008–2018. (3) The HXNR was found to be more effective than non-PA in improving ecosystem services. (4) The core zone of the nature reserve demonstrated better effectiveness in ecosystem service preservation.

Vegetation fluctuation is sensitive to climate change, and this response exhibits a time lag. Traditionally, scholars estimated this lag effect by considering the immediate prior lag (e.g., where vegetation in the current month is impacted by the climate in a certain prior month) or the lag accumulation (e.g., where vegetation in the current month is impacted by the last several months). The essence of these two methods is that vegetation growth is impacted by climate conditions in the prior period or several consecutive previous periods, which fails to consider the different impacts coming from each of those prior periods. Therefore, this study proposed a new approach, the weighted time-lag method, in detecting the lag effect of climate conditions coming from different prior periods. Essentially, the new method is a generalized extension of the lag-accumulation method. However, the new method detects how many prior periods need to be considered and, most importantly, the differentiated climate impact on vegetation growth in each of the determined prior periods. We tested the performance of the new method in the Loess Plateau by comparing various lag detection methods by using the linear model between the climate factors and the normalized difference vegetation index (NDVI). The case study confirmed four main findings: (1) the response of vegetation growth exhibits time lag to both precipitation and temperature; (2) there are apparent differences in the time lag effect detected by various methods, but the weighted time-lag method produced the highest determination coefficient (R²) in the linear model and provided the most specific lag pattern over the determined prior periods; (3) the vegetation growth is most sensitive to climate factors in the current month and the last month in the Loess Plateau but reflects a varied of responses to other prior months; and (4) the impact of temperature on vegetation growth is higher than that of precipitation. The new method provides a much more precise detection of the lag effect of climate change on vegetation growth and makes a smart decision about soil conservation and ecological restoration after severe climate events, such as long-lasting drought or flooding.

With the growing popularity of location-based social media applications, point-of-interest (POI) recommendation has become important in recent years. Several techniques, especially the collaborative filtering (CF), Markov chain (MC), and recurrent neural network (RNN) based methods, have been recently proposed for the POI recommendation service. However, CF-based methods and MC-based methods are ineffective to represent complicated interaction relations in the historical check-in sequences. Although recurrent neural networks (RNNs) and its variants have been successfully employed in POI recommendation, they depend on a hidden state of the entire past that cannot fully utilize parallel computation within a check-in sequence. To address these above limitations, we propose a spatiotemporal dilated convolutional generative network (ST-DCGN) for POI recommendation in this study. Firstly, inspired by the Google DeepMind’ WaveNet model, we introduce a simple but very effective dilated convolutional generative network as a solution to POI recommendation, which can efficiently model the user’s complicated short- and long-range check-in sequence by using a stack of dilated causal convolution layers and residual block structure. Then, we propose to acquire user’s spatial preference by modeling continuous geographical distances, and to capture user’s temporal preference by considering two types of time periodic patterns (i.e., hours in a day and days in a week). Moreover, we conducted an extensive performance evaluation using two large-scale real-world datasets, namely Foursquare and Instagram. Experimental results show that the proposed ST-DCGN model is well-suited for POI recommendation problems and can effectively learn dependencies in and between the check-in sequences. The proposed model attains state-of-the-art accuracy with less training time in the POI recommendation task.

Effectively identifying COVID-19 patients using non-PCR clinical data is critical for the optimal clinical outcomes. Currently, there is a lack of comprehensive understanding of various biomedical features and appropriate technical approaches to accurately detecting COVID-19 patients. In this study, we recruited 214 confirmed COVID-19 patients in non-severe (NS) and 148 in severe (S) clinical type, 198 non-infected healthy (H) participants and 129 non-COVID viral pneumonia (V) patients. The participants’ clinical information (23 features), lab testing results (10 features), and thoracic CT scans upon admission were acquired as three input feature modalities. To enable late fusion of multimodality data, we developed a deep learning model to extract a 10-feature high-level representation of the CT scans. Exploratory analyses showed substantial differences of all features among the four classes. Three machine learning models (k-nearest neighbor kNN, random forest RF, and support vector machine SVM) were developed based on the 43 features combined from all three modalities to differentiate four classes (NS, S, V, and H) at once. All three models had high accuracy to differentiate the overall four classes (95.4%-97.7%) and each individual class (90.6%-99.9%). Multimodal features provided substantial performance gain from using any single feature modality. Compared to existing binary classification benchmarks often focusing on single feature modality, this study provided a novel and effective breakthrough for clinical applications. Findings and the analytical workflow can be used as clinical decision support for current COVID-19 and other clinical applications with high-dimensional multimodal biomedical features. One sentence summary: We trained and validated late fusion deep learning-machine learning models to predict non-severe COVID-19, severe COVID-19, non-COVID viral infection, and healthy classes from clinical, lab testing, and CT scan features extracted from convolutional neural network and achieved predictive accuracy of > 96% to differentiate all four classes at once based on a large dataset of 689 participants.

This study mainly examined the relationships among primary productivity, precipitation and temperature by identifying trends of change embedded in time-series data. The paper also explores spatial variations of the relationship over four types of vegetation and across two precipitation zones in Inner Mongolia, China. Traditional analysis of vegetation response to climate change uses minimum, maximum, average or cumulative measurements; focuses on a whole region instead of fine-scale regional or ecological variations; or adopts generic analysis techniques. We innovatively integrate Empirical Mode Decomposition (EMD) and Redundancy Analysis (RDA) to overcome the weakness of traditional approaches. The EMD filtered trend surfaces reveal clear patterns of Enhanced Vegetation Index (EVI), precipitation, and temperature changes in both time and space. The filtered data decrease noises and cyclic fluctuations in the original data and are more suitable for examining linear relationship than the original data. RDA is further applied to reveal partial effect of precipitation and temperature, and their joint effect on primary productivity. The main findings are as follows: (1) We need to examine relationships between the trends of change of the variables of interest when investigating long-term relationships among them. (2) Long-term trend of change of precipitation or temperature can become a critical factor influencing primary productivity depending on local environments. (3) Synchronization (joint effect) of precipitation and temperature in growing season is critically important to primary productivity in the study area. (4) Partial and joint effects of precipitation and temperature on primary productivity vary over different precipitation zones and different types of vegetation. The method developed in this paper is applicable to ecosystem research in other regions.