Peter K. Kitanidis specializes in the analysis of data and the development and implementation of mathematical models that describe and predict flow and transport rates in the environment. He has devised methods for the analysis of spatially distributed hydrologic and water-quality data, data matching and uncertainty quantification of groundwater models, and the optimization of sampling and control strategies when the available information is incomplete.  He has developed methods for large-scale inverse modeling and data assimilation using fast linear algebra tools.  He has researched and developed novel inversion methods for hydraulic tomography (static, transient, and oscillatory).  He is active in the study of mixing processes in groundwater and the development of cost-effective enhanced in-situ remediation methods. He has been investigating how the heterogeneous nature of hydrogeochemical variables affects the net rates of flow and transport at hydrologic scales.  He is developing software for data analysis and monitoring at sites of geologic storage of CO2.  He has also developed methods for real-time forecasting of river flows.