Academic
Publications
GIST, A MODEL FOR GENERATING SPATIAL-TEMPORAL DAILY RAINFALL DATA

GIST, A MODEL FOR GENERATING SPATIAL-TEMPORAL DAILY RAINFALL DATA,Guillermo A. Baigorria,James W. Jones

GIST, A MODEL FOR GENERATING SPATIAL-TEMPORAL DAILY RAINFALL DATA  
BibTex | RIS | RefWorks Download
Weather generators are tools developed to create synthetic daily weather data over long periods of time. These tools have also been used for downscaling from monthly to seasonal forecasts produced by global and regional circulation models to daily values in order to provide inputs for crop and other environmental models. A major limitation of weather generators is that they do not take into account the spatial structure of weather and climate for a given region or watershed. This spatial correlation is important when one aggregates variables, for example, simulated crop yields or water resources, across a watershed or region. A method was developed to generate realizations of daily rainfall for multiple sites in an area while preserving the spatial and temporal patterns among sites. A two-step method generates rainfall events followed by rainfall amounts at sites where a generated rainfall event occurs. Generation of rainfall events was based on a two-state orthogonal Markov chain for discrete distributions. For generating rainfall amounts, a vector of random numbers ( ( ) 1 , 0 ~ N rnorm ) of order equal to the number of locations with rainfall events that were generated to occur in a specific day was matrix- multiplied by the corresponding reduced function of the correlation matrix to create correlated random numbers. To generate the final rainfall amounts, elements from the resulting vector of spatially-correlated random numbers were retransformed to a gamma distribution using cumulative probability functions calculated individually for each location. Values were next rescaled to rainfall amounts. Seven weather stations in North-Central Florida were selected, and a thousand replications of daily rainfall data were generated for this study. Rainfall events and amounts from the new method were compared to those from the WGEN point-based weather generator. The spatial structure in generated daily rainfall events and amounts closely matched the observed ones among all pairs of weather stations and other monthly rainfall statistics for each weather station. Correlation coefficient between observed and generated (ρo-g) joint probabilities that station pairs are both with rainfall was 0.996 and for both without rainfall was 0.991. The ρo-g correlation among weather stations was 0.983 for rainfall amounts and was significant at the 0.01 probability level. Root mean square errors of correlation values ranged from 0.04 to 0.08 for rainfall events and from 0.01 to 0.09 for amounts.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.