Search published articles


Showing 1 results for Multivariable Linear Regressions

A.a Sabziparvar, H Tabari, A Aeini,
Volume 14, Issue 52 (7-2010)
Abstract

Soil temperature is one of the important variables in hydrology, agriculture, meteorology and climatology studies. Owing to the fact that soil temperature is only measured at synoptic stations, reconstruction of this variable in other places is of great importance for many relevant agricultural surveys. Using 10-year (1996-2005) daily meteorological observations, including: air temperature, global solar radiation, precipitation, relative humidity, vapor pressure, wind speed and air pressure data, different empirical relationships are suggested. At statistically significant level (P<0.05), the suggested regressions are reliable for estimating soil temperature in various depths (5, 10, 20, 30, 50 and 100 cm) and different climate types. Using soil temperature as the dependent variable and the other meteorological parameters as the independent variables, the multivariable relationships are classified accordingly. The results indicate that the impact of meteorological parameters on soil temperature is not the same. At statistically significant level (P<0.05), the mean daily air temperature presented the highest correlation coefficients with soil temperature for all climate types (on average, from R2>0.91 for warm semi-arid, to R2>0.85 for humid climates). Other results highlighted that the correlation coefficients decreased as the soil depth increased. The behavior of statistical validation criteria of the suggested relations are also discussed for all the mentioned climates.

Page 1 from 1     

© 2024 CC BY-NC 4.0 | JWSS - Isfahan University of Technology

Designed & Developed by : Yektaweb