Establishment of the regression equation for water absorption in apple trees with auto flowing trunk injection

In this study, the interval estimation of the regression equation about the initial speed v0 and the absorption in estimation interval was determined. Pure water was used as the test material. Function relation of adjacent observation time corresponding to the flow speed was analyzed and regression equations of uptake with the initial speed v0 was established by statistical software. The results were as follows. The estimation interval of the regression equation was 0-48 h after injection. The adjacent observation time corresponding to the flow speed had significant linear and exponent curve relationship. Temperature had influence on the flow speed. When temperature rose, flow speed increased. When the temperature dropped, velocity also fell. Significant regression relationship was observed between the initial speed v0 and the absorption 48 h after injection. If the ratio of v0 and v3 was not distinguished, the difference rate of one-factor liner regression was significantly higher than exponent function regression, given that the value range of the initial speed v0 was between 0 mL/min and 1 mL/min. However, the difference rate of the exponent function regression was significantly higher than one-factor liner regression, when the value range of the initial speed v0 was between 2 mL/min and 4 mL/min. If the ratio of v0 and v3 was distinguished and the ratio range was between 0.6 and 1, the difference rate of one-factor liner regression was significantly lower than that of exponent function regression when the value range of v0 was between 1 mL/min and 4 mL/min. With the ratio range from 1 to 1.6, there was no significant difference among different rates, given the v0 value was between 0 mL/min and 4 mL/min. If only v0 was used to estimate the absorption, the range of different rates was between 20% and 30%. When the range of v0 was between 0 mL/min and 1 mL/min, exponent function regression was advised for the estimation of the absorption. When the range of v0 was between 1 mL/min and 2 mL/min, exponent function regression and one-factor liner regression were both able to estimating the absorption. When the range of v0 was between 2 mL/min and 4 mL/min, one-factor liner regression was advised for estimating the absorption. When using both v0 and the ratio of v0 and v3 to estimate the absorption, the value of difference rate was less than 15%. When the ratio range was between 0.6 and 1, exponent function regression and one-factor liner regression were both able to estimate the absorption with the range of v0 between 0 mL/min and 1 mL/min. One-factor liner regression was advised for estimating the absorption when the range of v0 was between 1 mL/min and 4 mL/min. When the ratio range was between 1 and 1.6, both the exponent function regression and one-factor liner regression were able to estimate the absorption with the range of v0 between 0 mL/min and 4 mL/min.