S4  Ciências e Engenharia do Petróleo
 
 Title:
SPATIAL CORRELATION IN DISCRETE MARKOV RANDOM FIELDS AND TRUNCATED GAUSSIAN PROCESSES
 
Summary :
THE CHARACTERIZATION OF PHYSICAL PROPERTIES OF PETROLEUM RESERVOIRS IS THE INITIAL POINT IN THE FORECASTING OF THE FLUIDS FLOW BEHAVIOR IN POROUS MEDIA. IN THIS SUBJECT, THE SIMULATION OF STOCHASTIC PROCESSES HAS AN IMPORTANT ROLE, BY GENERATING IMAGES OF THE VARIABLE, WITH SPECIFIC CHARACTERISTICS OF SPATIAL CORRELATION AND CONTINUITY. IN THIS WORK WE DISCUSS RELEVANT ASPECTS CONCERNING THE USE OF DISCRETE MARKOV RANDOM FIELDS AND GAUSSIAN STOCHASTIC PROCESSES IN THE REPRESENTATION OF ROCK PROPERTIES IN FLUIDS RESERVOIRS. THE MAIN POINT FOCUSED IS THE RELATIONSHIP BETWEEN THE SPATIAL CORRELATION IN GAUSSIAN PROCESSES AND THE ATTRACTION PARAMETER IN MARKOV RANDOM FIELDS, HERE STUDIED BY MEASURING THE AUTOCORRELATION PARAMETERS IN BINARY MARKOV IM-AGES, GENERATED BY THE METROPOLIS ALGORITHM. GAUSSIAN IMAGES WITH GAUSSIAN TYPE AUTOCORRELA-TION, AFTER TRUNCATED IN BINARY FACIES, HAS A CORRESPONDENCE TO MARKOV IMAGES. THIS SIMILARITY IS VALIDATED BY ANALYSIS IN THE AUTOCORRELATION FUNCTION OF THE DISCRETE GAUSSIAN PROCESSES.  
 
Author :
Remacre, Armando Zaupa
Salomão, Marcelo Curzio
 
 
Paper View :

 

COBEM99 | Organizing Committee | Keynotes | Technical  Sessions | Author | Symposia and special sessions | Referees
Work  Titles | Venue