Two‐dimensional shallow water models have been widely used in forecasting, risk assessment, and management of floods. Application of these models to large‐scale floods with high‐resolution terrain data significantly increases the computation cost. In order to reduce computation time, shallow water models are simplified by neglecting the inertial and/or convective acceleration terms in the momentum equations. The local inertial models have proved to significantly improve the computational efficiency even for large‐scale flood forecasting. However, instability issues are encountered on smooth surfaces of urban areas having low friction values. This problem was resolved by de Almeida et al. (2012, https://doi.org/10.1029/2011WR011570) by introducing limited artificial diffusion in the form of weighting factors for the neighboring fluxes. The arbitrary value of the weighting factor poses a practical limitation of being case specific and requiring calibration for accurate solutions. This study derives an explicit expression for the weighting factor, an adaptive formulation dependent on local velocity, flow depth, grid, and time step size, which eliminates the need for trials and approximations. Comparisons between analytical, experimental, and real‐world applications confirm the accuracy and robustness of the proposed weighting factor. Implementation of adaptive weights results in less computation time compared to LISFLOOD‐FP (~1.2 times) and holds a significant advantage over HEC‐RAS (~25.9 times) as it allows the use of larger time step at higher Courant‐Friedrichs‐Lewy (CFL) values. The contribution of the present study therefore resolves an important problem of current large‐scale flood simulations, especially those implemented in real time.

No abstract

No abstract

No abstract

Using a large Web search service as a case study, we highlight the challenges that modern Web services face in understanding and diagnosing the response time experienced by users. We show that search response time (SRT) varies widely over time and also exhibits counterintuitive behavior. It is actually higher during off-peak hours, when the query load is lower, than during peak hours. To resolve this paradox and explain SRT variations in general, we develop an analysis framework that separates systemic variations due to periodic changes in service usage and anomalous variations due to unanticipated events such as failures and denial-of-service attacks. We find that systemic SRT variations are primarily caused by systemic changes in aggregate network characteristics, nature of user queries, and browser types. For instance, one reason for higher SRTs during offpeak hours is that during those hours a greater fraction of queries come from slower, mainly-residential networks. We also develop a technique that, by factoring out the impact of such variations, robustly detects and diagnoses performance anomalies in SRT. Deployment experience shows that our technique detects three times more true (operator-verified) anomalies than existing techniques.