Urban heat Island to a significant degree high temperature phenomena some towns or in the confines of a village nearby called. Such a phenomenon has led to many problems. This study answers these questions by analyzing 130 years of temperature data of Mashhad. Grey System relational change-point method used to find the starting point of trend in meteorology and hydrology. Several parametric and non-parametric methods were used to linear trend in this study. The least-square tests, Mann-Kendall and Sense significant trend were confirmed after and before the change point. These tests are based on a comparison of null and one Hypotheses. The null hypothesis of this test is based on randomness and lack of trend data series and acceptance assume a (reject the null hypothesis) is evidence of trends in the data series. The three models of regression: polynomial, Soft Spline polynomials (Splines) and Multivariate adaptive regression splines (MARS) was fitted for analyzing trend non-linear of Mashhad temperature. The starting point of trend of Mashhad temperature (start heat island) Obtained 1979 by System relational change-point method .MARS and Spline regression analysis confirms that the warming trend in Mashhad is S-shaped and slope recent years has been very low after 2000 and relatively static state. This phenomenon may be due to the saturation of urban heat island and low trend after 2000 that can be caused by global warming. The results also showed that the average temperature in Mashhad after 1979 has increased by about 2.5 degrees.