Christopher Moore

A year without summer for California, or a harbinger of a climate shift?

Mark Twain once remarked “The coldest winter I ever spent was a summer in San Francisco.” If residents and visitors along the California coast think the summer of 1999 was even cooler than usual, they are right. Several air temperature records were set (R. Renard, pers. comm., 2000). October 1999 marked a 13-month run of below-normal temperatures at Monterey, and the March–July mean was 1.4°C below the long-term average.

GeoModeler: tightly linking spatially-explicit models and data with a GIS for analysis and geovisualization

Numerical simulation models provide a way to understand and predict the behavior of natural and human systems. Ideally, spatially-explicit models would be easily linked to a geographic information system (GIS) for data analysis and visualization. In the past, these two have not been well integrated for scientific uses. This lack of true integration hinders the ability of scientists and managers to create interactive, GIS-based models for research and policy planning. However, GIS packages are starting to expose code and objects to allow closer coupling of core GIS functionality and analytical/modeling tools. In creating GeoModeler, we have provided a prototype of how one might integrate a GIS with oceanographic and decision-support models. Through the use of Java-based application programming interfaces and connectors, a GIS is directly linked with the Regional Ocean Modeling System (ROMS) and with the Method of Splitting Tsunami (MOST) model. Scientists and managers are able to use a graphical interface to display datasets, select the data to be used in a scenario, set the weights for factors in the model and run the model. The results are returned to the GIS for display and spatial analysis. Three-dimensional visualizations are created using elements of the Visualization Toolkit (VTK) and OpenGL. The project creates a framework for linking to other types of back-end models written in a variety of programming languages.

The 2004 Sumatra tsunami in the Southeastern Pacific Ocean: New Global Insight from Observations and Modeling

The 2004 Sumatra tsunami was an unprecedented global disaster measured throughout the world oceans. The present study focused on a region of the southeastern Pacific Ocean where the “westward” circumferentially propagating tsunami branch converged with the “eastward” branch, based on data from fortuitously placed Chilean DART 32401 and tide gauges along the coast of South America. By comparison of the tsunami and background spectra, we suppressed the influence of topography and reconstructed coastal “spectral ratios” that were in close agreement with a ratio at DART 32401 and spectral ratios in other oceans. Findings indicate that even remote tsunami records carry spectral source signatures (“birth-marks”). The 2004 tsunami waves were found to occupy the broad frequency band of 0.25–10 cph with the prominent ratio peak at period of 40 min related to the southern fast-slip source domain. This rupture “hot-spot” of ∼350 km was responsible for the global impact of the 2004 tsunami. Data from DART 32401 provided validation of model results: the simulated maximum tsunami wave height of 2.25 cm was a conservative approximation to the measured height of 2.05 cm; the computed tsunami travel time of 25 h 35 min to DART 32401, although 20 min earlier than the actual travel time, provided a favorable result in comparison with 24 h 25 min estimated from classical kinematic theory. The numerical simulations consistently reproduced the wave height changes observed along the coast of South America, including local amplification of tsunami waves at the northern stations of Arica (72 cm) and Callao (67 cm).

Tsunami Simulation Using Sources Inferred from Various Measurement Data: Implications for the Model Forecast

Model forecast applications use various models of tsunami sources inferred from different measurement data. Even the same type of observation data can produce substantially different tsunami source models during a real-time forecast when more data are obtained during the real-time analysis. Improved tsunami observations enable investigation of the influence of such model source variability on the final forecast using different source data sets of several events. The 2010 Maule, Chile and 2011 Tohoku, Japan tsunamis were two recent events that provide ample observations throughout the Pacific and were, thus, used here to study the sensitivity of different model inputs for forecasting. The sources for these events were derived using the following three different methods: (1) real time or post event inversion of tsunameter water level data; (2) prediction of sea floor deformations via analysis of seismic wave forms and application of a finite fault model; and (3) prediction of sea floor deformation using real-time GPS data. For the March 11, 2011 Tohoku tsunami, two examples of each method are used, while for the February 27, 2010 Maule event, only one tsunameter inversion and one finite fault model method were used due to a much more limited data set. Observed data from the Deep-ocean Assessment and Reporting for Tsunamis (DART) network, Japan GPS buoys, and select tide gauges across the Pacific were compared with forecasts to assess the sensitivity of these three methods using root-mean-square error analysis. We divided the analysis by the type of data and the distance from the source. This sensitivity analysis showed that increasing the resolution of a tsunami source model does not necessarily improve tsunami forecast quality, even in the near-field. Instead, the findings suggest that when forecasting coastal impact, defining the overall energy characteristic of a tsunami source may be more important than refining small source details. Source models based on direct tsunami observations are better at reproducing a tsunami signal: this finding is not very surprising but has implications for tsunami forecasting and warning operations.

Integration of Tsunami Analysis Tools into a GIS Workspace – Research, Modeling, and Hazard Mitigation efforts Within NOAA’s Center for Tsunami Research

The National Oceanic and Atmospheric Administration’s (NOAA) Center for Tsunami Research (NCTR) uses geospatial data and GIS analysis techniques in support of building an accurate tsunami forecasting system for the US Tsunami Warning Centers. The resulting forecast products can be integrated into applications and visualizations to assess hazard risk and provide mitigation for US coastal communities ranging from small towns to large urban centers. NCTR also conducts basic research on the nature of tsunami propagation and inundation, which relies on accurate geospatial information. In this chapter, we discuss how we have used both open source and commercially available geospatial technologies to address issues in tsunami research and hazard mitigation – including model visualization, data delivery, and emergency management products. Additionally, we discuss the development and coupling of tsunami model results with coastal risk, vulnerability, and evacuation models, raising the issues of integration, visualization, proliferation of mapping applications, and the ease of use and intended audience of these products.