PROGRAM Kagan_2009_GJI_I_scores A program for scoring earthquake-forecast grids by the methods of: Kagan, Yan Y. [2009] Testing long-term earthquake forecasts: likelihood methods and error diagrams, Geophys. J. Int., v. 177, pages 532-542. Some advantages of these methods are that they: -are insensitive to the grid used to cover the Earth; -are insensitive to changes in the overall seismicity rate; -do not modify the locations or magnitudes of test earthquakes; -do not require simulation of virtual catalogs; -return relative quality measures, not just "pass" or "fail;" and -indicate relative specificity of forecasts as well as relative success. This program by Peter Bird, UCLA, May 2013 & March 2014. Masking added October 2014. Converted to REAL*8 and 64-bit in July 2015. ------------------------------------------------------------------------- REQUIRED INPUT: AT LEAST ONE FORECAST: *A stationary forecast of seismicity rate with a single magnitude bin (everything at or exceeding threshold) presented at gridded longitude/latitude points, according to the Long_Term_Seismicity variant of the .grd (Gridded Data) format documented at: http://peterbird.name/guide/grd_format.htm (NOTE that the Long_Term_Seismicity variant of the .grd format inserts the threshold magnitude into the first line, and also adds 2 lines with the forecast integral after the end of the long number list.) Enter name of forecast .GRD file: SHIFT_GSRM2e_5p767_2005plus.grd Here are the first 2 lines of this file: -179.87500 0.25000 179.87500 = lon_min, d_lon, lon_max Gridded long-term seismicity above magnitude 5.767 in events/square-me -89.80000 0.20000 89.80000 = lat_min, d_lat, lat_max from program SHIFT_GSRM2x, P. Bird, UCLA, 2014 What is the threshold magnitude of the forecast? 5.76700000000000 What is the maximum depth (in km, using INTEGER) for this forecast? 70 *Optionally, a "mask.grd" file (in Peter Bird's .GRD format) can be read to block the use of certain forecast cells in computations of both I_0 and I_1. An entry of "1" in this .grd file indicates masking; an entry of "0" indicates normal processing of that cell. Do you wish to read in a MASK.GRD file now?: F ------------------------------------------------- I_0 = 3.9144 (specificity of forecast) ------------------------------------------------- REQUIRED INPUT: EARTHQUAKE CATALOG for the test time-window. *A seismic catalog for (exactly) the test time-window, which is either global or spatially larger than the forecast region; which extends at least as deep as the depth limit of forecast(s); which is complete down to the threshold magnitude of the forecast(s). The catalog must be in Peter Bird's .eqc = EarthQuake Catalog format (as output by program Seismicity) which is documented in: Bird, P., and Y. Y. Kagan [2004] Plate-tectonic analysis of shallow seismicity: Apparent boundary width, beta, corner magnitude, coupled lithosphere thickness, and coupling in seven tectonic settings, Bull. Seismol. Soc. Am., 94(6), 2380-2399, plus electronic supplement (eqc_format.pdf). NOTE that the .EQC catalog file should be trimmed to the exact time window of the intended tests. NOTE that the optional focal-mechanism information at the end of each line in the .eqc file is not currently used. Enter name of EQC file: GCMT_shallow_m5.767_2014-2018.eqc Do you want an output file listing score per test EQ?: T Writing KaganIs_per-EQ.dat... Computing forecast success (I_1)... 905 relevant earthquakes found in test catalog. ------------------------------------------------- I_1 = 3.6613 (success of forecast) ------------------------------------------------- Work completed.