Geostatistical surfaces of climatological normals of mean air temperature in Minas Gerais.

Air temperature, a vital component for the terrestrial environment sustainability, can be used as an indicator and an important factor used in short- and long-term meteorological modeling at different scales. Temperature must be monitored on spatial and temporal scale with high precision. Terrain elevation can be used as the main influence factor depending on the measurement scale. In small and medium scales, factors related to local relief were modeled with geostatistics including external variables in temperature modeling. We aimed to evaluate the use of universal kriging in the modeling of air temperature in order to create temperature surfaces at each km[Formula: see text] in Minas Gerais State, Brazil using altitude, longitude and latitude covariates. The organized mean air temperature data of climatological normals of the National Institute of Meteorology were submitted to summary statistics, statistical regression and geostatistical analysis. Monthly and annual normals of the mean air temperature compensated for the period 1981 to 2010 were modeled using temperature as dependent variable and altitude, longitude and latitude as co-variables. Multiple regression modeling performed on temperature using altitude, longitude and latitude covariates determined significant parameters for monthly and annual mean air temperature global prediction. Relief and coordinates were used as external drift on variography and universal kriging with block for local temperature interpolation and prediction in order to generate 1-km moderate resolution surfaces of monthly and annual mean air temperature. Universal kriging determined smoothing effect of standard deviation of geospatial variation with prediction errors varying between 0.6 and [Formula: see text]C. Higher prediction error values were observed between June and August. Mean air temperature local prediction presented greater errors mainly in the lower altitude regions and in the colder months. In both monthly and annual temperature predictions, universal kriging with external drift enabled to circumvent the problem of performing spatial prediction from sparse punctual attribute data, conferring a temperature downscaling effect in Minas Gerais.

Human progress and drought sensitivity behavior.

OBJECTIVES: Human progress is the satisfaction of human needs associated with life quality, social and economic justice. Considering that climate had significant effect on socio-economic behavior of Earth populations, the objective of this work was to evaluate the relationship between human progress and drought sensitivity behavior in Minas Gerais State, Brazil. METHODS: The human progress variable was generated by principal component analysis, considering the factors: Territorial occupation of the population, dependency ratio, urbanization degree, income, habitation, health, education, sanitation, security and human development index. The moisture index product was generated based on map algebra calculation in GIS, using WorldClim precipitation and MODIS evapotranspiration products. Nonlinear least squares method and Levenberg-Marquardt algorithm fit an exponential model to the moisture index (independent) and human progress (dependent) variables. Covariance and crossed-covariance functions were used to characterize the structure and magnitude of spatial dependence between the human progress and the moisture index co-variable. RESULTS: Couple modeling human progress and moisture index variability generated the human drought sensitivity behavior variable, representing the effects of regional climate in Homo sapiens population. The generated human drought sensitivity behavior values were inverted, i.e. lower and higher human drought sensitivity behavior values indicated higher and lower human sensitivity to drought, respectively. Adaptive management capacity with more capable governments combining economic and engineering solutions to maximize species, habitat, ecosystem survival and resilience, could generate best mitigation solutions to reduce potential impacts of climatic spatial variation in Minas Gerais state, Brazil.