Integrated Hydrologic Modeling to Untangle the Impacts of Water Management During Drought.

Over the past century, groundwater levels in California's San Joaquin Valley have dropped by more than 30 m in some areas mostly due to excessive groundwater extraction used to irrigate agricultural lands and sustain a growing population. Between 2012 and 2015, California experienced the worst drought in its recorded history, depleting surface water supplies and further exacerbating groundwater depletion in the region. Due to a lack of groundwater regulation, exact quantities of extracted groundwater in California are unknown and hard to quantify. Recent adoption of the Sustainable Groundwater Management Act has intensified efforts to identify sustainable groundwater use. However, understanding sustainable use in a highly productive agricultural system with an extremely complex surface water allocation system, variable groundwater use, and spatially extensive and diverse irrigation practices is no easy task. Using an integrated hydrologic model coupled with a land surface model, we evaluated how water management activities, specifically a suite of irrigation and groundwater pumping scenarios, impact surface water-groundwater fluxes and storage components and how those activities and the relationships between them change during drought. Results showed that groundwater pumping volume had the most significant impact on long-term water storage changes. A comparison with total water storage anomaly (TWSA) estimates from NASA's Gravity Recover and Climate Experiment (GRACE) provided some insight regarding which combinations of pumping and irrigation matched the GRACE TWSA estimates, lending credibility to these scenarios. In addition, the majority of long-term water storage changes during the recent drought occurred in groundwater storage in the deeper subsurface.

Restoring Speech Following Total Removal of the Larynx.

By speech articulator movement and training a transformation to audio we can restore the power of speech to someone who has lost their larynx. We sense changes in magnetic field caused by movements of small magnets attached to the lips and tongue. The sensor transformation uses recurrent neural networks.

Restoring speech following total removal of the larynx by a learned transformation from sensor data to acoustics.

Total removal of the larynx may be required to treat laryngeal cancer: speech is lost. This article shows that it may be possible to restore speech by sensing movement of the remaining speech articulators and use machine learning algorithms to derive a transformation to convert this sensor data into an acoustic signal. The resulting "silent speech," which may be delivered in real time, is intelligible and sounds natural. The identity of the speaker is recognisable. The sensing technique involves attaching small, unobtrusive magnets to the lips and tongue and monitoring changes in the magnetic field induced by their movement.