Analysis of SJ’s climatology, 2010 and 2011-Rain vs Temp

[ROCKstormSJ4315]

Extraido del tópico: Estaciones Meteorologicas

Informacion de precipitacion y temperatura
Postby chelo_00 » Sun Oct 11, 2015 10:13 pm

Saludos,

Tremenda pagina primera ves que entro y tiene muy buena informacion. Nesecito una ayuda con un proyecto ( no vi la informacion en su pagina) es de "La Relacion enttre la temperatura y la cantidad de precipitacion en PR en los meses de Agosto a Octubre".No he encontrado mucha informacion y no se donde buscarla. Cualquier link o informacion que me puedan brindar se lo agradesco.

Muchas Gracias

Chelo(Cayey)
chelo_00
Onda Tropical
Onda Tropical

Posts: 1
Joined: Sun Oct 11, 2015 9:58 pm

Chelo, tal como te indique en el otro topico, aqui colocando una parte de uno de varios estudios que prepare como parte de mi curso de SAS y que seleccione como tema la meteorologia. El estudio original tiene mas de 100 paginas y habia puesto mas informacion en la pagina del chat anterior que ya no existe. Lo que estoy poniendo es un "draft" por lo que puede tener varios errores gramaticales. Mas adelante tratare de colocar mas informacion.



Analysis of San Juan’s climatology during 2010 and 2011.


Regression Models- Multivariate Model

Table 16 The REG Procedure
Model: MODEL1
Dependent Variable: DICWTR: Dichotomous 0=Rain, 1= Not rain

Number of Observations Read 730
Number of Observations Used 730


Analysis of Variance

Sum of Mean
Source DF Squares Square F Value Pr > F

Model 9 51.52360 5.72484 32.74 <.0001
Error 720 125.87914 0.17483
Corrected Total 729 177.40274


Root MSE 0.41813 R-Square 0.2904
Dependent Mean 0.58356 Adj R-Sq 0.2816
Coeff Var 71.65127


Parameter Estimates

Parameter Standard
Variable Label DF Estimate Error t Value Pr > |t|

Intercept Intercept 1 -0.20295 0.50943 -0.40 0.6905
MAX Max Temp. for Day 1 -0.01231 0.00794 -1.55 0.1215
MIN Min Temp for Day 1 0.01157 0.00791 1.46 0.1441
WTR Precipitation Amount 1 0.19371 0.03229 6.00 <.0001
AVGSPD Average Wind Speed 1 -0.03269 0.00742 -4.40 <.0001
MAXSPD The highest wind Speed over 2 1 0.02111 0.00994 2.12 0.0340
Minutes
DIR Wind Direction in Degree 1 0.00012000 0.00033873 0.35 0.7232
Format
SS The Average Sky Cover during 1 0.11368 0.01288 8.83 <.0001
day
SPD Peak wind Speed 1 0.01159 0.00819 1.41 0.1578
DR Peak wind Direction for day in 1 -0.00059759 0.00030400 -1.97 0.0497
Degrres





The Table 16 shows the variance which can be explained by the nine independent variables Model (9) and the variance which is not explained by the independent variables or the Sum of Square Error (720). As reflected the coefficient of determination, the nine variables included in the model do not explain a significant portion of the variation of rain for the season(29%). The second part of the table shows the coefficient of each variables. The most important thing on this table is to interpret the coefficients and how significant are they in the behavior of rain. With the results of table 8 hypothesis test is possible to determine if the coefficients (β) are significant. In the table, marked in red, shows that only 3 of the nine variables are significant to explain if one day it will rain or not going to rain. Those variables are Average Wind Speed, The Highest Wind Speed and the Average Sky Cover.

Ho: β1 = β2 = β3 = β4 = 0 if p-value is greater than alpha the null hypothesis is correct.

P-values for many coefficients on the table are greater than the value of alpha; therefore the null hypothesis is correct. Some of the coefficients for independent variables are not statistically significantly different from 0. From the three variables, the most crucial variable to explain the rainfall is cloud cover. It is assumed that if it is cloudy there is good chance of rain. However, as the correlation of variables with the dependent variable is so small, it is assumed that not necessarily because it is cloudy, there will be a rain event in Puerto Rico.

In conclusion, although three of the nine variables are significant variables for the model projection, do not provide enough to explain whether it will rain in Puerto Rico.

It is noted that the same thing I ever said in other "post", in relation to sampling errors, may be present here. . I used for this analysis SAS Statistics.

However, regardless of the mistakes that can have this analysis, which is done by an amateur, that the study does not have all the scientific rigor, the results showed relevant information with reference to rain and temperatures in San Juan.

Final note: Analysis done by an amateur. Do not be taken as official information. Refer to NOAA for information or official studies.

[YankeeStorm]

SAS Statistics, great folks!

[ROCKstormSJ4315]

SAS Statistics, great folks!

Gracias amigo. Aqui entrando al topico hoy despues de casi un año. Cuando dispongo de tiempo aplico algo de estadistica con algunos de esos software interesantes para manejo de datos.