Search

Ghost forests and abandoned farms are stark indicators of ecological change along world coastlines, caused by sea level rise (SLR). These changes adversely affect terrestrial ecosystems and economies, but expanding coastal marshes resulting from SLR also provide crucial ecosystem services such as carbon sequestration and mediate material fluxes to the ocean. We introduce a US-NSF Critical Zone Network project designed to untangle the hydrological, ecological, geomorphological, and biogeochemical processes that are altering the functioning of the marsh-upland transition in the coastal critical zone. We have instrumented six sites in the mid-Atlantic region of the US, along the coastlines of the Atlantic Ocean, Delaware Bay, and Chesapeake Bay where marshes are rapidly encroaching into forests and farmland. We have installed field sensors to observe the effects of slow hydrologic change (i.e. SLR) and fast episodic events such as high tides and storm surges on water levels, land surface elevation, salinity, redox conditions, and sap flow. We are coupling these measurements to laboratory experiments and analyses, as well as modeling to elucidate drivers and feedbacks in these complex and highly transient critical zone systems.

Interpreting time domain reflectometry (TDR) waveforms obtained in soils with non-uniform water content is an open question. We design a new TDR waveform interpretation model based on convolutional neural networks (CNNs) that can reveal the spatial variations of soil relative permittivity and water content along a TDR sensor. The proposed model, namely TDR-CNN, is constructed with three modules. First, the geometrical features of the TDR waveforms are extracted with a simplified version of VGG16 network. Second, the reflection positions in a TDR waveform are traced using a 1D version of the region proposal network. Finally, the soil relative permittivity values are estimated via a CNN regression network. The three modules are developed in Python using Google TensorFlow and Keras API, and then stacked together to formulate the TDR-CNN architecture. Each module is trained separately, and data transfer among the modules can be facilitated automatically. TDR-CNN is evaluated using simulated TDR waveforms with varying relative permittivity but under a relatively stable soil electrical conductivity, and the accuracy and stability of the TDR-CNN are shown. TDR measurements from a water infiltration study provide an application for TDR-CNN and a comparison between TDR-CNN and an inverse model. The proposed TDR-CNN model is simple to implement, and modules in TDR-CNN can be updated or fine-tuned individually with new datasets. In conclusion, TDR-CNN presents a model architecture that can be used to interpret TDR waveforms obtained in soil with a heterogeneous water content distribution., One popular way to measure water content in a whole soil profile is to insert an array of short TDR sensors at multiple, pre-specified depths. Despite the efforts needed to install the sensors and the potential disturbance of nearby soils, it provides discretized measurements in space. Instead, vertically inserting long TDR sensors could provide measurement without the restriction of pre-specific depths, but the spatial variation of soil water content along the TDR sensor rods needs to be reconstructed. That is an open and challenging question in TDR data interpretation studies. We developed a deep machine learning model that can compute the soil water content distributions along the sensor rods, and the model performance is evaluated using simulated TDR data with varying soil water content but under relatively stable electrical conductivity. The artificial neural network architecture in this study presents a model prototype, which is important for scientists and engineers in TDR data analysis, sensor design and related industries.