RAPIDLY GROWING VOLUMES OF INSAR IMAGERY FROM CURRENT (SENTINEL ALOS-2) AND FORTHCOMING (NISAR) SATELLITES MISSIONS PRESENT NEW OPPORTUNITIES TO CHARACTERIZE GEOLOGIC AND ANTHROPOGENIC PROCESSES THAT DRIVE ACTIVE DEFORMATION. AT THE SAME TIME THE SHEER VOLUME OF OBSERVATIONS PRESENTS DATA MINING CHALLENGES IN STRAIN DETECTION QUALITY CONTROL AND TIME SERIES ANALYSIS. I PROPOSE TO DEVELOP AND APPLY ARTIFICIAL NEURAL NETWORKS (ANNS) TO DETECT AND CLASSIFY STRAIN IN INSAR TIME SERIES TO DECONVOLVE INSAR TIME SERIES INTO ITS CONTRIBUTING DEFORMATION AND NOISE SIGNALS AND TO GENERATE HIGH-PRECISION STRAIN RATE MEASUREMENTS OVER THE ARABIA-EURASIA PLATE BOUNDARY. ANNS ARE OPTIMIZED TO CLASSIFY LARGE SPATIAL AND TEMPORAL DATA SETS AND TO WORK WELL WITH NON-LINEAR OBSERVATIONS. ADDITIONALLY ANNS ARE TRAINED WITH WELL-DOCUMENTED MODELS AND THEY REQUIRE NO PRIOR ASSUMPTIONS CONCERNING THE NATURE OF THE OBSERVATIONS. TO TRAIN DEBUG AND IMPLEMENT THE ANN I WILL TAKE THE FOLLOWING APPROACHES WITH THREE DIFFERENT ANN METHODOLOGIES: TEST ANN CAPABILITIES WITH SYNTHETIC NOISY AND NOISE-FREE INSAR OBSERVATIONS; TEST ANN CAPABILITIES WITH KNOWN WELLDOCUMENTED SURFACE DISPLACEMENT SIGNALS; IMPLEMENT ANN APPROACHES ON A CONTINENTAL-SCALE INSAR TIME SERIES. I WILL GENERATE A CONTINENTAL-SCALE INSAR TIME SERIES OVER THE ARABIA-EURASIA PLATE BOUNDARY OF IRAQ IRAN AND PAKISTAN AN IDEAL NATURAL LABORATORY FOR INSAR TECHNIQUE DEVELOPMENT GIVEN THE FREQUENCY AND DIVERSITY OF SURFACE DEFORMATION PROCESSES. IN THE PROPOSED STUDY REGION THIS WORK WILL ALLOW ME TO ADDRESS FUNDAMENTAL QUESTIONS WITH RESPECT TO THE ACCOMMODATION OF ACTIVE CONVERGENCE BETWEEN ARABIA/ INDIA AND EURASIA: WHERE IS STRAIN CURRENTLY ACCOMMODATED AND AT WHAT RATES WHERE ARE THERE POTENTIALLY UNIDENTIFIED SEISMIC HAZARDS AND HOW DOES TRANSIENT STRAIN DETECTABILITY VARY THROUGHOUT THE STUDY AREA? THROUGH THIS RESEARCH I WILL DEVELOP ANN APPROACHES TO INSAR TIME SERIES ANALYSIS WHICH WILL IMPROVE COMMUNITY CAPABILITIES TO CLASSIFY SURFACE DISPLACEMENTS AND DECONVOLVE INSAR TIME SERIES INTO CONTRIBUTING SIGNALS (DEFORMATION AND NOISE)