Moldova_Atlas
eng
MD_CharacterSetCode_utf8
dataset
Institute of Ecology and Geography Academy of Sciences, Moldova
marianedealcov@yahoo.com
pointOfContact
2014-11-19
ISO19115
2003/Cor.1:2006
Atlas. Climatic resources of the Republic of Moldova
Moldova_Atlas
The atlas contains a set of thematic maps reflecting the spatial distribution of monthly, seasonal, and annual mean air temperature, the average amount of monthly, seasonal and annual precipitations for a period of 30 years (1981-2010). The data recorded from meteorological stations and posts of State Hidrometeorological Service served as starting material. All maps were developed at scale 1:1500000 in Universal Transversal Mercator projection (UTM), using cartographic modelling. Collected data corresponds to meteorological shelter height (2 m).
Institute of Ecology and Geography Academy of Sciences, Moldova
marianedealcov@yahoo.com
author
climate
Drought Vocabulary
2011-04-09
publication
No conditions apply
otherRestrictions
no limitation
50000
200
eng
climatologyMeteorologyAtmosphere
26.62
30.17
45.47
48.49
1981-01-01
2010-12-31
unknown
unknown
http://ieg.asm.md/sites/default/files/Atlas-2013.pdf
dataset
Commission Regulation (EC) No 1205/2008 of 3 December 2008
2008-12-04
publication
See the referenced specification
Maps was performed in two stages. At the first step was used multiple regression with step procedures, that allowed the highlighting the values, that reflecting the temperature and rainfall dependence of several local physical-geographical factors. As an indicator of the validation of the models were used: the physical-geographical significance of each factor taken separately and for the entire model, the coefficient of determination R2, the standard error of estimate SEE, and the mean absolute error MAE. The P value is smaller, the confidence level CL is increased. P value <= 0.1 corresponds to CL = 90%. For P <= 0.05, CL = 95%, and for P <= 0.01, CL = 99%. F is the ratio of the variance explained by regression to the unexplained. The coefficient of determination R2 is the percentage of variability in the dependent variable explained by the regression. SEE is equal to the standard deviation of residues (differences between baseline values and those obtained by regression) and MAE – the average residuals.At the second stage regression residuals were interpolated (which are determined by unknown factors) using a local interpolator. The results of the interpolations were summed with the results of the regression model. The interpolators take into account only local neighborhood data for interpolated point.In the case of temperature as interpolator was used procedure IDW (Inverse Distance Weighted). The influence of neighboring points decreases with distance from interpolated point. The function exponent (usually equal to 2) was optimized.The precipitation has a dynamic character and a high variability in space. Therefore, we used a larger set of data, and as a local interpolator, the residual Kriging was used, that take into account the spatial structure of the data. In this case, it was necessary that the residuals satisfy certain conditions. At first, the residuals should be normally distributed and the average should be equal to 0. The standardized coefficients of skewness and kurtosis needs to be placed in the range of +- 2. The statistical analysis shows that these conditions are fulfilled in most cases. In case when the values of standardized residuals were outside the range +- 2, these extreme values (outliers) were removed from the calculation (some precipitation stations). At the same time, was takes into account the spatial trend, which in most cases was approximated by a polygon of second order. Another test method used was the cross validation, which stipulate the exclusion of subsequent data and comparing the results with the original data. Statistical analysis of data was performed in Statgraphics Centurion XV program and cartographic modeling and development of the final product – in the ArcGIS 10.