\n<\/strong><\/p>\n
BAT is designed and developed by Dr. Ramesh Teegavarapu with the help of Dr. A. Goly to evaluate radar data using several visual, quantitative and statistical indices. It is can also be used for evaluation of hydrologic models and any simulation model-generated results. BAT can be used for forecast verification of precipitation forecasts available from National Weather Service (NWS). BAT consists of several error and performance metrics for evaluation of radar data at different temporal and spatial resolutions. BAT can be customized to client-specific requirements and expanded with additional features. BAT will help to answer these questions:<\/strong><\/p>\n What is the overall quality of the radar data compared to rain gage data (i.e., ground truth)?<\/span><\/em> BAT provides a comprehensive evaluation using:<\/strong><\/p>\n Four association measures<\/span> Modules: The main modules in BAT are:<\/span><\/p>\n 1. Data Input<\/span> Visual<\/span><\/span><\/p>\n BAT consists of multiple visual assessment measures for assessment of data. Quantitative<\/span><\/span><\/p>\n Statistical<\/span><\/span><\/p>\n Interactive Analysis<\/span><\/span><\/p>\n Comparative Analysis<\/span><\/p>\n A few graphical results from BAT<\/span><\/p>\n Variogram clouds provide details about the spatial dependence.<\/span><\/p>\n Multi-model\/product comparisons in one diagram with three error statistics<\/span><\/p>\n _________________<\/p>\n <\/p>\n","protected":false},"excerpt":{"rendered":" _________________________ Bias Analysis Tool (BAT) BAT is designed and developed by Dr. Ramesh Teegavarapu with the help of Dr. A. Goly to evaluate radar data using several visual, quantitative and statistical indices. It is can also be used for evaluation … Continue reading
\n<\/span><\/p>\n
\nAre the radar and rainfall data characteristics statistically similar?<\/span><\/em>
\nWhat is the overall bias?<\/span><\/em>
\nWhat is the error structure and how it is varying in space and time?<\/span><\/em>
\nWhat is variability of the bias in space and time?<\/span><\/em>
\nWhat is the skill of the method used for radar-based rainfall estimation in the definition of rain or no rain events?<\/span><\/em>
\nDo the radar-based rainfall estimates preserve the site (at a rain gage location) and regional (a set of rain gages) statistics?<\/span><\/em>
\nAre the rainfall fields generated by radar-based estimates and rain gage-based observations similar?<\/span><\/em>
\nCan a specific index or a set of indices help provide directions to improve the radar-based rainfall data?<\/span><\/em>
\nHow do radar and rain gage data quantitatively compare at different temporal resolutions?<\/span><\/em>
\nIs radar data quality appropriate for distributed hydrologic modeling?<\/span><\/em><\/p>\n
\nSix quantitative error performance measures<\/span>
\nSix skill scores<\/span>
\nSix contingency measures<\/span>
\nSix summary statistics<\/span>
\nThree statistical tests<\/span>
\nSix Proximity metrics<\/span>
\nMulti-model\/product evaluations<\/span>
\nUncertainty Assessment for performance measures<\/span>
\nAssessment at multiple spatial resolutions<\/span><\/p>\n
\n<\/strong><\/p>\n
\n A module to accept data in a specified format<\/span>
\n 2. Data Preparation<\/span>
\n This module prepares the data set for analysis and check the data format and completeness<\/span>
\n 3. Visual<\/span>
\n This is a non-interactive module for evaluation of visual indices<\/span>
\n 4. Quantitative<\/span>
\n This is a non-interactive module for evaluation of quantitative indices<\/span>
\n 5. Statistical<\/span>
\n This is a non-interactive module for evaluation of statistical indices<\/span>
\n 6. Interactive Analysis<\/span>
\n This is an interactive module as the name suggests for all the three (visual, quantitative and statistical indices)<\/span>
\n 7. Comparative Analysis<\/span>
\n This is again an interactive module with option of comparative analysis option (side-by-side) for all indices for two different temporal time frames.<\/span>
\n 8. Report<\/span>
\n Generates two types of reports: Tabular and Graphical for the data selected in the \u201cData Preparation\u201d module.<\/span><\/p>\n
\n<\/span><\/p>\nScatter plots, Box plots, Quantile-Quantile plots, Accumulated Value plots,\u00a0 Cumulative Distribution Function Plots, Kernel Density Plots, Residual Plots, Transitional Probability Plots.<\/span><\/h4>\n
![\"Bat-visual1\"](\"http:\/\/faculty.eng.fau.edu\/ramesh\/files\/2015\/03\/Bat-visual1.png\")
<\/a><\/p>\nError Measure: Mean Error (ME), Mean Absolute Error (MAE), Multiplicative Bias (MB), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Linear Error in Probability Space (LEPS), Correlation Coefficient (CC), Nash-Sutcliffe Efficiency Coefficient (NSCE), Hiedke Skill Score (HSS), Bootstrap sampling-based Error Measure (BSEM), <\/span><\/h4>\n
Contingency Measures: Sensitivity, Error Rate, Specificity, Concordance<\/span><\/h4>\n
<\/a><\/p>\nSummary Statistics: Mean, Standard Deviation, Minimum value, Maximum value, Skewness, Kurtosis. <\/span>Check for Distributions: Kolmogorov-Smirnov (two sample test), Ansari Bradley Test, and Wilcoxon Test<\/span><\/h4>\n
<\/a><\/p>\n<\/a><\/p>\nIn interactive analysis mode, BAT can be used by the user to select visual, quantitative and statistical metrics to evaluate data at a site or multiple sites.<\/span><\/h5>\n
<\/a><\/p>\nIn comparative analysis mode, BAT can be used by the user to select two temporal windows for assessment of data for single or multiple sites.
\n<\/span><\/h4>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\nFor more details please contact Dr. T.\u00a0 at rteegava@fau.edu<\/a><\/span><\/h4>\n