This study aims at estimating sum of global irradiation amounts at location (lat. 40.07.16 North and long 43.35.00 East) of migratory birds found in Iğdır province of Turkey. In the estimation of global irradiation parameters (Hd: average daily sum of global irradiation per square meter and Hm: average annually sum of global irradiation per square meter), several predictors viz. ESTLOSTEMP (estimated losses due to temperature and low irradiance), ESTLOSANGREF (estimated loss due to angular reflectance effect), and COMPVLOSS (Combined Photo Voltaic system losses) were calculated. Estimation of global irradiation parameters was made through multivariate adaptive regression splines (MARS) data mining algorithm for multiple responses (Hd and Hm) with the support of R software program and the utility prediction equation was aimed to improve for further biodiversity investigations. To determine the predictive quality of the MARS algorithm, goodness of fit criteria viz. coefficient of determination (0.994 and 0.996 R2 for Hd and Hm), Generalized Cross Validation (0.000038 and 0.024024 GCV for Hd and Hm), Cross-Validation R2 (0.974 and 0.967 CVR2 for Hd and Hm), Residual Sum of Squares (0.00046 and 0.28829 RSS for Hd and Hm) and Standard Deviation Ratio (0.078 and 0.063 SDRATIO for Hd and Hm) were calculated for penalty= -1 in the package “earth” of the R software. MARS prediction equation was derived at the smallest estimates of GCV which is defined as the ratio of RSS to n (sample size) for penalty= -1. The smallest GCV values were set at number of terms (4). Goodness of fit criteria exhibited that the MARS prediction model had a very good fit for a cross validation of 3. As a result, the obtained results may be baseline information about global irradiation parameters at location of migratory birds in Iğdir, Turkey for next studies with the scope of global warming.