%{search_type} search results

Relative crustal motions along active faults generate earthquakes, and repeated motions build mountain ranges over millions of years. However, the long-term summation of elastic, earthquake-related deformation often cannot produce the deformation recorded within the rock record. Here, we provide an explanation for this discrepancy by showing that increases in strain facilitated by plastic deformation of Earth’s crust, in conjunction with isostatic deflection and erosion, transform relative fault motions that produce individual earthquakes to geologic deformations. We focus our study on the data-rich Santa Cruz Mountains, CA, USA, and compare predicted and observed quantities for rock uplift, apatite (U-Th)/He thermochronology, topographic relief, 10Be-based erosion rates, and interseismic surface velocities. This approach reconciles these disparate records of mountain-building processes spanning spatial scales from millimeters to tens of kilometers, allowing us to explicitly bridge decadal measures of deformation with that produced by millions of years of plate motion. This repository contains code, data files, and visualization products associated with this work.

In this code supplement to the paper "LOCA: LOcal Conformal Autoencoder for standardized data coordinates" we offer a Python software library that includes: - Implementation of the LOCA algorithm - Scripts that generate each of the figures in this paper.

We derive a formula for optimal hard thresholding of the singular value decomposition in the presence of correlated additive noise; although it nominally involves unobservables, we show how to apply it even where the noise covariance structure is not a-priori known or is not independently estimable. The proposed method, which we call ScreeNOT, is a mathematically solid alternative to Cattell's ever-popular but vague Scree Plot heuristic from 1966. ScreeNOT has a surprising oracle property: it typically achieves exactly, in large finite samples, the lowest possible MSE for matrix recovery, on each given problem instance - i.e. the specific threshold it selects gives exactly the smallest achievable MSE loss among all possible threshold choices for that noisy dataset and that unknown underlying true low rank model. The method is computationally efficient and robust against perturbations of the underlying covariance structure. Our results depend on the assumption that the singular values of the noise have a limiting empirical distribution of compact support; this model, which is standard in random matrix theory, is satisfied by many models exhibiting either cross-row correlation structure or cross-column correlation structure, and also by many situations where there is inter-element correlation structure. Simulations demonstrate the effectiveness of the method even at moderate matrix sizes. The paper is supplemented by ready-to-use software packages implementing the proposed algorithm.

Example code and spreadsheets to illustrate the calculations from the paper "Efficient analytical fragility function fitting using dynamic structural analysis."

This repository contains the codes and synthesized data for our paper "Learning from past respiratory infections to predict COVID-19 Outcomes: A retrospective study". In this paper we proposed a framework that used COVID-like cohorts (you can find the description in our paper) to train machine learning models and validated them on the COVID-19 population.If you just want to test the model on synthetic COVID dataset, you could run the predict_COVID.py directly, or if you want to train the model by your own data, you could run the "training_models.py" file.The file descriptions are:directory: contains the test data. is the data template we used.directory: contains the trained models by COVID-like patients.is used to predict intubation within 48 hours.is used to load test data from test data file.is used to train the modelsProcess:1) Input the command "pip install -r requirements.txt" to install the environment2) Using python command to train the model: python training_models.py3) Using python command to run the model: python predict_COVID.pyThe output is a score file in the "result" directory.

Accurate estimates of groundwater storage changes are critical to the sustainable management of worldwide aquifers, but few methods exist that can assess both local and regional aquifer dynamics, especially in regions where groundwater monitoring wells are sparse. In this study, we developed a water mass balance methodology that integrates remotely sensed and in situ hydrologic data and models to estimate changes in groundwater storage at multiple spatial scales. To demonstrate this method, we quantified monthly groundwater storage changes in California’s Central Valley Alluvial Aquifer and the surrounding watershed from 2001 to 2019. Groundwater storage changes are calculated as the residual of inflows (precipitation, streamflow, runoff) minus outflows (evapotranspiration, streamflow), corrected for fluctuations in surface water storage (reservoirs, snow) and soil moisture. Trends, timing, and magnitude of our results agree with independent estimates of groundwater storage change calculated from (a) water levels measured in groundwater wells, (b) the Gravity Recovery and Climate Experiment (GRACE) (c) regional groundwater flow models, and (d) estimates derived from GPS. The results reproduce periods of drought (2011-2015) and the exceptionally wet 2016-2017 California winter. Uncertainty in estimates of groundwater storage changes corresponding to (a-d) was quantified through a triple collocation analysis. The use of remote sensing data to monitor sub-regional to regional changes in groundwater storage can contribute significantly to the sustainable management of aquifers, particularly in parts of the world with limited access to monitoring well data, hydrogeologic information, and groundwater flow models.

Accurate and timely estimates of groundwater storage changes are critical to the sustainable management of worldwide aquifers, but few methods exist that can assess both local and regional aquifer dynamics, especially in regions where groundwater monitoring wells are sparse. In this study, we developed a water mass balance methodology that uses primarily remotely sensed data and models to estimate changes in groundwater storage at multiple spatial scales. To demonstrate this method, we quantified monthly groundwater storage changes in California’s Central Valley and the surrounding watershed from 2001 to 2019. Groundwater storage changes are calculated as the residual of inflows (precipitation, streamflow, runoff) minus outflows (evapotranspiration, streamflow), corrected for fluctuations in surface water storage (reservoirs, snow) and soil moisture. Trends, timing, and magnitude of our results agree with independent estimates of groundwater storage change calculated from (a) water levels measured in groundwater wells, (b) the Gravity Recovery and Climate Experiment (GRACE) (c) regional groundwater flow models, and (d) estimates derived from GPS. The results reproduce periods of drought (2011-2015) and the exceptionally wet 2016-2017 California winter. Uncertainty in estimates of groundwater storage changes corresponding to (a-d) was quantified through ensemble models and triple collocation analyses. The use of remote sensing data to monitor sub-regional to regional changes in groundwater storage can contribute significantly to the sustainable management of aquifers, particularly in parts of the world with limited access to monitoring well data, hydrogeologic information, and groundwater flow models.