|
Registro Completo |
|
Biblioteca(s): |
Embrapa Pantanal. |
|
Data corrente: |
09/03/1999 |
|
Data da última atualização: |
17/12/2008 |
|
Autoria: |
BRUHN, J. A.; FRY, W. E.; FICK, G. W. |
|
Afiliação: |
Department of Plant Pathology, Cornell University, Ithaca. Department of Agronomy, Cornell University. |
|
Título: |
Simulation of daily weather data using theoretical probability distributions. |
|
Ano de publicação: |
1980 |
|
Fonte/Imprenta: |
Journal of Applied Meteorology, v.19, n.9, p.1029-1036, Sept. 1980. |
|
Idioma: |
Inglês |
|
Conteúdo: |
A computer simulation model was constructed to supply daily weather data to a plant disease management model for potato late blinght. In the weather model Monte Carlo techniques were employed to generate daily values of precipitation, maximum temperature, minimum temperature, minimum relative humidity and total and total solar radiation. Each weather variable is described by a known theoretical probability distribution but the values of the parameters describing each distribution are dependent on the occurrence of rainfall. Precipitation occurrence is described by a first-order Markov chain. The amount of rain, given that rain has occurred, is described by a gamma probability distribution. Maximum and minimum temperature are simulated with a trivariate normal probability distribution involving maximum temperature on the previous day, maximum temperature on the current day and minimum temperature on the current day. Parameter values for this distribution are dependent on the occurrence of rain on the previous day. Both minimum relative humidity and total solar radiation are aassumed to be normally distributed. The values of the parameters describing the distribution of minimum relative humidity is dependent on rainfall occurrence on the previous day and current day. Parameter values for total solar radiation are dependent on the occurrence of rain on the current day. The assumptions made during model construction were found to be appropriate for actual weather data from Geneva, New York. The performance of the weather model was evaluated by comparing the cumulative frequency distributions of simulated weather data with the distribution of actual weather data from Geneva, New York and Fort Collins, Colorado. For each location, simulated weather data were similar to actual weather data in terms of mean response, variability and autocorrelation. The possible applications of this model when used with models of other components of the agro-ecosystem are discussed. MenosA computer simulation model was constructed to supply daily weather data to a plant disease management model for potato late blinght. In the weather model Monte Carlo techniques were employed to generate daily values of precipitation, maximum temperature, minimum temperature, minimum relative humidity and total and total solar radiation. Each weather variable is described by a known theoretical probability distribution but the values of the parameters describing each distribution are dependent on the occurrence of rainfall. Precipitation occurrence is described by a first-order Markov chain. The amount of rain, given that rain has occurred, is described by a gamma probability distribution. Maximum and minimum temperature are simulated with a trivariate normal probability distribution involving maximum temperature on the previous day, maximum temperature on the current day and minimum temperature on the current day. Parameter values for this distribution are dependent on the occurrence of rain on the previous day. Both minimum relative humidity and total solar radiation are aassumed to be normally distributed. The values of the parameters describing the distribution of minimum relative humidity is dependent on rainfall occurrence on the previous day and current day. Parameter values for total solar radiation are dependent on the occurrence of rain on the current day. The assumptions made during model construction were found to be appropriate for actual weather data from Genev... Mostrar Tudo |
|
Palavras-Chave: |
Metorology. |
|
Thesagro: |
Agricultura; Meteorologia. |
|
Thesaurus Nal: |
agriculture. |
|
Categoria do assunto: |
-- |
|
Marc: |
LEADER 02526naa a2200193 a 4500
001 1794052
005 2008-12-17
008 1980 bl --- 0-- u #d
100 1 $aBRUHN, J. A.
245 $aSimulation of daily weather data using theoretical probability distributions.
260 $c1980
520 $aA computer simulation model was constructed to supply daily weather data to a plant disease management model for potato late blinght. In the weather model Monte Carlo techniques were employed to generate daily values of precipitation, maximum temperature, minimum temperature, minimum relative humidity and total and total solar radiation. Each weather variable is described by a known theoretical probability distribution but the values of the parameters describing each distribution are dependent on the occurrence of rainfall. Precipitation occurrence is described by a first-order Markov chain. The amount of rain, given that rain has occurred, is described by a gamma probability distribution. Maximum and minimum temperature are simulated with a trivariate normal probability distribution involving maximum temperature on the previous day, maximum temperature on the current day and minimum temperature on the current day. Parameter values for this distribution are dependent on the occurrence of rain on the previous day. Both minimum relative humidity and total solar radiation are aassumed to be normally distributed. The values of the parameters describing the distribution of minimum relative humidity is dependent on rainfall occurrence on the previous day and current day. Parameter values for total solar radiation are dependent on the occurrence of rain on the current day. The assumptions made during model construction were found to be appropriate for actual weather data from Geneva, New York. The performance of the weather model was evaluated by comparing the cumulative frequency distributions of simulated weather data with the distribution of actual weather data from Geneva, New York and Fort Collins, Colorado. For each location, simulated weather data were similar to actual weather data in terms of mean response, variability and autocorrelation. The possible applications of this model when used with models of other components of the agro-ecosystem are discussed.
650 $aagriculture
650 $aAgricultura
650 $aMeteorologia
653 $aMetorology
700 1 $aFRY, W. E.
700 1 $aFICK, G. W.
773 $tJournal of Applied Meteorology$gv.19, n.9, p.1029-1036, Sept. 1980.
Download
Esconder MarcMostrar Marc Completo |
|
Registro original: |
Embrapa Pantanal (CPAP) |
|
