This supplementary material provides the MATLAB application that is introduced in the main article. In addition, we present some figures and tables for better understanding of the application of the multifractal seismicity analysis method. Performing fractal analysis on large datasets using computer code often can be a difficult task for seismologists. Thus, we have assembled the important multifractal analysis codes under a graphical user interface (GUI) named Fractal Analyzer for easy handling. This tool is capable of performing multifractal seismicity analysis for large datasets. In addition, it has some graph-plotting utilities that supplement the main application and help avoid the need for different graph-plotting software. In the main article, we demonstrated the applicability of this application by using it to perform fractal analysis on the 2005 Kashmir, India, earthquake event.
The application and the user manual are zipped together in FractalAnalyzer.zip. On unzipping the file, two different files will be extracted: FractalAnalyzer.bac and FractalAnalyzer.pdf. The first step is to change the. bac extension of the application to. exe. (Because some of web uploaders do not allow uploading of executable files, we changed the file extension to. bac.) Once, the executable file is ready, we recommend the user read FractalAnalyzer.pdf on the requirements and procedure to run the application.
Download: FractalAnalyzer.zip [zipped archive file; ~3 MB]. This zipped archive file consists of two files:
Figure S1. Screenshot of FractalAnalyzer application in Multifractal Computation mode.
Figure S2. Screenshot of FractalAnalyzer application in Grapher mode.
Figure S3. Screenshot of FractalAnalyzer application in log Cq(r) versus log r is shown for one time window of the Kashmir earthquake region to find Dq with R2. (Cq(r) is the correlation integral; r is distance; Dq is a generalized dimension; q is an order of generalized dimension.)
Figure S4. Screenshot of FractalAnalyzer application in Dq – q plot or Dq spectrum for a window with q = 2 to 22. As the q value increases, the exponential decay of the Dq value shows the multifractal nature for the events.
Figure S5. Screenshot of FractalAnalyzer application in the temporal variation of multifractal dimension Dq (for q = 2 or Dc, which is the Dq value computed for each cluster of seismic events belonging to the study region) for the spatial distribution of seismic events (mb ≥ 3.5).
Figure S6. Grapher display of FractalAnalyzer application in the temporal variation of Dc for the spatial distribution of seismic events with error (mb ≥ 3.5) prior to the Kashmir earthquake on 8 October 2005 (Mw 7.7) as indicated in Table S1.
Figure S7. Grapher display of FractalAnalyzer application in the temporal variation of Dc for the spatial distribution of seismic events with error (mb ≥ 3.5) after the Kashmir earthquake on 8 October 2005 (Mw 7.7) as indicated in Table S2.
Figure S8. Spatial distribution for the latitude (32° N–37° N) and longitude (72° E–77° E) of two time windows of 50 events each with two low Dc values (recent events with mean time window 2005.083 years indicated by red color and the preceding events with mean time window 2003.25 years in blue) estimated as the precursor for the main shock of 8 October 2005 (Mw 7.7) indicated in Table S1. The location of the Kashmir epicenter is indicated with the star. MBT, Main Boundary thrust.
Table S1. Correlation dimension (Dc) value estimated with time for the events lying in the region of latitude (32° N–37° N) and longitude (72° E–77° E) prior to main shock with the estimation error and R2 as plotted in Figure S6.
Table S2. Correlation dimension (Dc) value estimated with time for the events lying in the region of latitude (32° N–37° N) and longitude (72° E–77° E) after the main shock with the estimation error and R2 as plotted in Figure S7.
[ Back ]