A5091206679- Soil Geostatistics and Mapping
- Geochemistry and Geologic Mapping
- Soil and Unsaturated Flow
- Reservoir Engineering and Simulation Methods
- Soil Management and Crop Yield
- Geological Modeling and Analysis
- Mycorrhizal Fungi and Plant Interactions
- Statistical and numerical algorithms
- Bayesian Modeling and Causal Inference
- Mineral Processing and Grinding
- Data Management and Algorithms
- Hydraulic Fracturing and Reservoir Analysis
- Geotechnical Engineering and Soil Stabilization
- Water resources management and optimization
- Hydrocarbon exploration and reservoir analysis
- Image Processing and 3D Reconstruction
- Plant Reproductive Biology
- Plant Molecular Biology Research
- Advanced Statistical Process Monitoring
- Economic, financial, and policy analysis
- Neural Networks and Applications
- Gaussian Processes and Bayesian Inference
- Statistical Mechanics and Entropy
- Geophysical and Geoelectrical Methods
- Data Analysis with R
Institut National de la Recherche Agronomique du Niger
2009
UCLouvain
2001-2006
Brazilian Agricultural Research Corporation
2004
We address the problem of prediction a spatial categorical variable by revisiting maximum entropy approach. first argue that, for predicting category probabilities, approach is more natural than least‐squares approach, such as (co‐)kriging indicator functions. then show knowing categories observed at surrounding locations, conditional probability observing location obtained with particular principle simple combination sums and products univariate bivariate probabilities. This equation can be...
Current soil process models require the most accurate values for each of their input parameters at finest spatial scale. Traditionally, property are obtained either from maps or geostatistical methods using exact laboratory measurements. Both data types convey substantial information: provide exhaustive but soft (vague) information, whereas analyses hard (accurate) scarce Ideally, they should be combined. This objective can reached a recently developed method, namely Bayesian maximum entropy...
Summary Categorical variables such as water table status are often predicted using the indicator kriging (IK) formalism. However, this method is known to suffer from important limitations that most frequently solved by ad hoc solutions and approximations. Recently, Bayesian Maximum Entropy (BME) approach has proved its ability predict categorical efficiently in a flexible way. In paper, we apply Ooypolder data set for prediction of classes sample set. BME compared with IK global well local...
Summary Pluri-Gaussian Simulation is now widely used for simulating facies properties. One of its advantages to be able reproduce transitions between lithotypes in the geological domain by mean a truncation diagram or lithotype rule. However, context non-stationary proportions remains difficult because most implementations lack coherent and automatic algorithm modify so that it matches local proportions. The proposed method aims at proposing solution this limitation. An adapted version CART...
In geostatistical facies models, the sedimentary genetic processes are often neglected. As a consequence, realizations not very realistic in terms heterogeneity patterns. this paper, following methodology initiated by Hu and al. 1994, random simulation process is developed for lobe sand bodies modelling. It used to produce model of lobes internal architecture fill space between horizons deposition. This encouraging training images that can be good candidates multiple point geostatistics but...
Abstract Two main methods are used to provide robust production forecasts on existing and new wells. The first option relies building a geomodel which reservoir simulations performed that require long engineering time. second is use the Decline Curve Analysis (DCA) type curve construction. This method delivers fast accurate results but solely data, neglecting geology, mechanisms often not enough precise predict future wells’ performance. In this paper, two workflows combining geostatistics...
Summary This paper is presenting a comparison of two methods to quantify uncertainties on production profiles for unconventional reservoirs: the traditional method with type curves fitting and statistical analysis geostatistical time series decomposition interpolation. Both are applied true case study several hundred wells data base. The results in favor approach which provides very promising validations. some way forwards suggested.