(Luận văn) nghiên cứu đặc điểm cấu trúc và sinh trưởng của rừng trồng keo lai (acacia auriculiformis acacia mangium) trên các nhóm đất trồng khác nhau ở khu vực huyện định quán, tỉnh đồng nai

BỘ GIÁO DỤC VÀ ĐÀO TẠO BỘ NÔNG NGHIỆP VÀ PTNT TRƯỜNG ĐẠI HỌC LÂM NGHIỆP TRẦN MINH HOÀNG lu an va n NGHIÊN CỨU ĐẶC ĐIỂM CẤU TRÚC VÀ to gh tn SINH TRƯỞNG CỦA RỪNG TRỒNG KEO LAI p ie (Acacia auriculiformis* Acacia mangium) w TRÊN CÁC NHÓM ĐẤT TRỒNG KHÁC NHAU Ở d oa nl KHU VỰC HUYỆN ĐỊNH QUÁN, TỈNH ĐỒNG NAI ll u nf va an lu oi m z at nh LUẬN VĂN THẠC SỸ LÂM NGHIỆP z m co l gm @ an Lu Đồng Nai, năm 2017 n va ac th si BỘ GIÁO DỤC VÀ ĐÀO TẠO BỘ NÔNG NGHIỆP VÀ PTNT TRƯỜNG ĐẠI HỌC LÂM NGHIỆP TRẦN MINH HOÀNG NGHIÊN CỨU ĐẶC ĐIỂM CẤU TRÚC VÀ lu an SINH TRƯỞNG CỦA RỪNG TRỒNG KEO LAI va n (Acacia auriculiformis* Acacia mangium) KHU VỰCHUYỆN ĐỊNH QUÁN, TỈNH ĐỒNG NAI p ie gh tn to TRÊN CÁC NHÓM ĐẤT TRỒNG KHÁC NHAU Ở nl w CHUYÊN NGÀNH: LÂM HỌC d oa MÃ SỐ: 60.62.02.01 u nf va an lu ll LUẬN VĂN THẠC SỸ LÂM NGHIỆP oi m z at nh z NGƯỜI HƯỚNG DẪN KHOA HỌC: PGS TS PHẠM THẾ DŨNG m co l gm @ an Lu Đồng Nai, năm 2017 n va ac th si CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM Độc lập – Tự – Hạnh phúc LỜI CAM ĐOAN Tôi cam đoan, cơng trình nghiên cứu riêng Các số liệu, kết nêu luận văn trung thực chƣa đƣợc công bố cơng trình nghiên cứu khác Nếu nội dung nghiên cứu trùng lặp với cơng trình nghiên lu cứu cơng bố, tơi xin hồn tồn chịu trách nhiệm tn thủ kết luận an đánh giá luận văn Hội đồng khoa học va n Đồng Nai, ngày 05tháng năm 2017 to p ie gh tn Người cam đoan d oa nl w Trần Minh Hoàng ll u nf va an lu oi m z at nh z m co l gm @ an Lu n va ac th i si LỜI CẢM ƠN Sau thời gian đƣợc học tập, tiếp thu kiến thức chuyên môn theo chƣơng trình đào tạo Thạc sỹ chuyên ngành Lâm học Trƣờng Đại học Lâm nghiệp, đến khóa học kết thúc Đƣợc cho phép Ban giám hiệu Trƣờng Đại học Lâm nghiệp, thực đề tài: “Nghiên cứu đặc điểm cấu trúc sinh trƣởng rừng trồng Keo lai (Acacia auriculiformis*Acacia mangium) nhóm đất trồng khác khu vực huyện Định Quán, tỉnh Đồng Nai”làm luận văn lu thạc sỹ khoa học lâm nghiệp an Trong trình thực đề tài, ngồi nỗ lực thân, va n nhận đƣợc hƣớng dẫn tận tình thầy giáo hƣớng dẫn PGS.TS Phạm Thế gh tn to Dũng, quan tâm, hỗ trợ Ban giám hiệu,cán Ban Khoa học công ie nghệ, thầy cô giáo củaTrƣờng Đại học Lâm nghiệp; giúp đỡ nhiệt tình p cán bộ, cơng nhân viên Công ty TNHH MTV Lâm nghiệp La Ngà – Đồng nl w Nai bạn bè đồng nghiệp trình thu thập, xử lý số liệu d oa Để hoàn thành đề tài luận văn này, cho phép tơi đƣợc bày tỏ lịng biết an lu ơn sâu sắc đến Thầy giáo hƣớng dẫn PGS.TS Phạm Thế Dũng, Ban giám u nf va hiệu nhà trƣờng, cán Ban Khoa học công nghệ, thầy cô giáo giảng viên trực tiếp truyền thụ kiến thức suốt khóa học Cơ sở 2- Trƣờng ll oi m Đại học Lâm nghiệp;cán bộ, công nhân viên Công ty TNHH MTV Lâm z at nh nghiệp La Ngà – Đồng Nai bạn bè đồng nghiệp đãgiúp đỡ Luận văn kết nghiên cứu tôi, số liệu kết z nghiên cứu luận văn hoàn toàn trung thực chƣa đƣợc sử dụng @ l gm để bảo vệ học vị Các tài liệu tham khảo có nguồn trích dẫn rõ ràng m co Đồng Nai, ngày 05 tháng năm 2017 an Lu ac th ii n va Trần Minh Hoàng si MỤC LỤC Trang LỜI CAM ĐOAN i LỜI CẢM ƠN ii MỤC LỤC iii NHỮNG KÝ HIỆU VÀ CHỮ VIẾT TẮT TRONG LUẬN VĂN v DANH SÁCH CÁC BẢNG vii DANH SÁCH CÁC HÌNH viii ĐẶT VẤN ĐỀ lu an Chƣơng n va TỔNG QUAN tn to 1.1 Khái quát Keo lai 1.2 Tình hình nghiên cứu cấu trúc rừng gh p ie 1.2.1 Khái niệm cấu trúc rừng 1.2.2 Tình hình nghiên cứu cấu trúc rừng giới nl w 1.2.3 Tình hình nghiên cứu cấu trúc rừng Việt Nam d oa 1.3 Tình hình nghiên cứu sinh trƣởng rừng an lu 1.3.1 Khái niệm sinh trƣởng rừng va 1.3.2 Tình hình nghiên cứu sinh trƣởng rừng giới ll u nf 1.3.3 Tình hình nghiên cứu sinh trƣởng rừng Việt Nam 13 oi m 1.4 Những nghiên cứu rừng trồng Keo lai 17 z at nh 1.5 Thảo luận chung 19 Chƣơng 21 z MỤC TIÊU, ĐỐI TƢỢNG, PHẠM VI, NỘI DUNG 21 gm @ VÀ PHƢƠNG PHÁP NGHIÊN CỨU 21 l 2.1 Mục tiêu nghiên cứu 21 m co 2.1.1 Mục tiêu tổng quát 21 an Lu 2.1.2 Mục tiêu cụ thể 21 2.2 Đối tƣợng nghiên cứu 21 ac th iii n va 2.3 Phạm vi nghiên cứu 21 si 2.4 Nội dung nghiên cứu 22 2.5 Phƣơng pháp nghiên cứu 22 2.5.1 Phƣơng pháp luận 22 2.5.2 Phƣơng pháp thu thập số liệu 22 2.5.3 Phƣơng pháp xử lý số liệu 24 2.5.4 Cơng cụ tính tốn 27 Chƣơng 28 ĐẶC ĐIỂM TỰ NHIÊN KHU VỰC NGHIÊN CỨU 28 3.1 Vị trí địa lý 28 lu 3.2 Địa hình 29 an n va 3.3 Khí hậu - Thủy văn 30 Chƣơng 33 gh tn to 3.4 Tài nguyên đất 31 p ie KẾT QUẢ NGHIÊN CỨU VÀ THẢO LUẬN 33 4.1 Cấu trúc lâm phần rừng trồng Keo lai 33 nl w 4.1.1 Cấu trúc đƣờng kính lâm phần 33 d oa 4.1.2 Cấu trúc chiều cao lâm phần 40 an lu 4.2.2 Sinh trƣởng chiều cao 52 va 4.3 Khảo sát số nhân tố ảnh hƣởng đến sinh trƣởng rừng Keo lai u nf trồng khu vực huyện Định Quán 57 ll 4.3.1 Ảnh hƣởng tuổi rừng 57 m oi 4.3.2 Ảnh hƣởng nhóm đất trồng 61 z at nh 4.4 Sinh trƣởng thể tích Keo lai 65 z 4.4.1 Sinh trƣởng thể tích Keo lai trồng nhóm đất đỏ vàng 65 gm @ 4.4.2 Sinh trƣởng thể tích Keo lai trồng nhóm đất đen 67 l KẾT LUẬN VÀ KIẾN NGHỊ 71 m co Kết luận 71 Kiến nghị 72 an Lu TÀI LIỆU THAM KHẢO 73 ac th iv n va PHỤ LỤC 78 si NHỮNG KÝ HIỆU VÀ CHỮ VIẾT TẮT TRONG LUẬN VĂN Ký hiệu, Nội dung, ký hiệu chữ viết tắt chữ viết tắt Tuổi rừng tuổi rừng, năm Cv Hệ số biến động D1.3 Đƣờng kính thân vị trí 1,3 m D1.3lt Đƣờng kính thân lý thuyết vị trí 1,3 m Dbq Đƣờng kính thân bình qn vị trí 1,3 m Dmax Đƣờng kính thân lớn vị trí 1,3 m Dmin Đƣờng kính thân nhỏ vị trí 1,3 m Ex Độ nhọn Hbq Chiều cao vút bình quân lu A an n va p ie gh tn to Chiều cao vút lớn Hmin Chiều cao vút nhỏ oa Chiều cao vút d Hvn nl w Hmax lu Chiều cao vút lý thuyết Ln Logarit số e N Số Pd Suất tăng trƣởng đƣờng kính vị trí 1,3 m Pdlt Suất tăng trƣởng đƣờng kính lý thuyết vị trí 1,3 m Ph Suất tăng trƣởng chiều cao Phlt Suất tăng trƣởng chiều cao lý thuyết ll u nf va an Hvnlt oi m z at nh z Sk Độ lệch Sx Sai tiêu chuẩn an Lu Phƣơng sai m co S2 l Biên độ biến động gm @ R n va ac th v si Lƣợng tăng trƣởng thƣờng xuyên hàng năm đƣờng kính Zd Lƣợng tăng trƣởng thƣờng xuyên hàng năm lý thuyết Zdlt đƣờng kính Lƣợng tăng trƣởng thƣờng xuyên hàng năm chiều cao Zh Lƣợng tăng trƣởng thƣờng xuyên hàng năm lý thuyết Zhlt chiều cao Lƣợng tăng trƣởng bình quân năm đƣờng kính ∆dlt Lƣợng tăng trƣởng bình qn năm lý thuyết đƣờng kính ∆h Lƣợng tăng trƣởng bình qn năm chiều cao ∆hlt Lƣợng tăng trƣởng bình quân năm lý thuyết chiều cao lu ∆d an n va p ie gh tn to d oa nl w ll u nf va an lu oi m z at nh z m co l gm @ an Lu n va ac th vi si DANH SÁCH CÁC BẢNG Bảng 4.1 Đặc trƣng phân bố N/D1.3 33 Bảng 4.2 Đặc trƣng phân bố N/Hvn 40 Bảng 4.3 Sinh trƣởng D1.3 lâm phần Keo lai trồng nhóm đất đỏ vàng 48 Bảng 4.4 Các dạng phƣơng trình biểu thị quy luật tƣơng quan đƣờng kính tuổi Keo lai (D1.3/A) trồng nhóm đất đỏ vàng sau thử lu nghiệm 49 an va Bảng 4.5 Sinh trƣởng D1.3 lâm phần Keo lai trồng nhóm đất đen 50 n Bảng 4.6 Các dạng phƣơng trình biểu thị quy luật tƣơng quan 51 gh tn to đƣờng kính tuổi Keo lai (D1.3/A) 51 ie trồng nhóm đất đen sau thử nghiệm 51 p Bảng 4.7 Sinh trƣởng Hvn lâm phần Keo lai trồng nhóm đất đỏ nl w vàng 53 d oa Bảng 4.8 Các dạng phƣơng trình biểu thị quy luật tƣơng quan chiều cao an lu tuổi Keo lai (Hvn/A) trồng nhóm đất đỏ vàng sau thử u nf va nghiệm 53 Bảng 4.9 Sinh trƣởng Hvn lâm phần Keo lai trồng nhóm đất đen 55 ll oi m Bảng 4.10.Các dạng phƣơng trình biểu thị quy luật tƣơng quan chiều cao z at nh tuổi Keo lai (Hvn/A) trồng nhóm đất đỏ vàng sau thử nghiệm 56 z Bảng 4.11 Sinh trƣởng D1.3 (cm) lâm phần Keo lai khu vực nghiên cứu 62 @ l gm Bảng 4.12 Sinh trƣởng Hvn (m) lâm phần Keo lai khu vực nghiên cứu 64 m co Bảng 4.13 Sinh trƣởng thể tích Keo lai trồng nhóm đất đỏ vàng 65 Bảng 4.14.Các dạng phƣơng trình biểu thị quy luật tƣơng quan thể tích an Lu tuổi Keo lai (V/A) trồng nhóm đất đỏ vàng sau thử nghiệm ac th vii n va 65 si Bảng 4.15 Sinh trƣởng thể tích Keo lai trồng nhóm đất đen 67 Bảng 4.16 Các dạng phƣơng trình biểu thị quy luật tƣơng quan thể tích tuổi Keo lai (V/A) trồng nhóm đất đen sau thử nghiệm 67 Bảng 4.17 Sinh trƣởng thể tích lâm phần Keo lai trồng nhóm đất khu vực nghiên cứu 69 lu an n va p ie gh tn to d oa nl w ll u nf va an lu oi m z at nh z m co l gm @ an Lu n va ac th viii si The mean absolute error (MAE) of 0.0621168 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is less than 0.05, there is an indication of possible serial correlation at the 95.0% confidence level Plot the residuals versus row order to see if there is any pattern that can be seen Simple Regression - D Den vs A Dependent variable: D Den (Duong Kinh Den) Independent variable: A (Tuoi) Squared-Y model: Y = sqrt(a + b*X) Coefficients Least Squares Estimate -58.6186 44.9939 lu Parameter Intercept Slope an n va p ie gh tn to Analysis of Variance Source Sum of Squares Model 56684.7 Residual 699.261 Total (Corr.) 57384.0 Standard Error 9.99472 2.23489 Df T Statistic -5.86496 20.1325 Mean Square 56684.7 139.852 P-Value 0.0020 0.0000 F-Ratio 405.32 P-Value 0.0000 d oa nl w Correlation Coefficient = 0.993889 R-squared = 98.7814 percent R-squared (adjusted for d.f.) = 98.5377 percent Standard Error of Est = 11.8259 Mean absolute error = 8.46592 Durbin-Watson statistic = 1.45745 (P=0.0736) Lag residual autocorrelation = -0.0350393 u nf va an lu ll The StatAdvisor The output shows the results of fitting a squared-Y model to describe the relationship between D Den and A The equation of the fitted model is oi m z at nh D Den = sqrt(-58.6186 + 44.9939*A) z gm @ Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between D Den and A at the 95.0% confidence level m co l The R-Squared statistic indicates that the model as fitted explains 98.7814% of the variability in D Den after transforming to a reciprocal scale to linearize the model The correlation coefficient equals 0.993889, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 11.8259 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu an Lu n va ac th OO si The mean absolute error (MAE) of 8.46592 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level lu an n va p ie gh tn to d oa nl w ll u nf va an lu oi m z at nh z m co l gm @ an Lu n va ac th PP si Phụ lục KẾT QUẢ TƯƠNG QUAN GIỮA HVN VÀ A TRÊN NỀN ĐẤT ĐỎ VÀNG TỪ CÁC HÀM MẶC ĐỊNH TRONG STATGRAPHICS PLUS CENTURION XV.I Comparison of Alternative Models - H Do Vang vs A lu an n va p ie gh tn to d oa nl w R-Squared 99.86% 99.67% 99.56% 98.93% 98.49% 98.12% 98.03% 97.60% 97.05% 97.01% 95.80% 92.41% 92.02% 90.51% 85.84% 85.45% 84.18% 77.37% 69.77% 66.95% 64.87% 46.14% ll u nf va an lu Correlation 0.9993 0.9983 0.9978 0.9947 0.9924 0.9906 -0.9901 0.9879 0.9851 0.9849 0.9788 0.9613 0.9593 0.9514 0.9265 -0.9244 0.9175 0.8796 -0.8353 0.8182 -0.8054 -0.6792 m oi z at nh Model Square root-Y logarithmic-X Square root-X Squared-Y Double reciprocal Logarithmic-X Multiplicative S-curve model Double square root Squared-Y square root-X Linear Double squared Logarithmic-Y square root-X Square root-Y Squared-Y logarithmic-X Squared-X Reciprocal-X Exponential Square root-Y squared-X Squared-Y reciprocal-X Logarithmic-Y squared-X Reciprocal-Y Reciprocal-Y squared-X Reciprocal-Y square root-X Reciprocal-Y logarithmic-X Square root-Y reciprocal-X Logistic Log probit z @ m co l gm The StatAdvisor This table shows the results of fitting several curvilinear models to the data Of the models fitted, the square root-Y logarithmic-X model yields the highest R-Squared value with 99.864% This is 2.85721% higher than the currently selected linear model To change models, select the Analysis Options dialog box Simple Regression - H Do Vang vs A Dependent variable: H Do Vang (CCao Do vang) Independent variable: A (Tuoi) an Lu n va ac th QQ si Square root-Y logarithmic-X model: Y = (a + b*ln(X))^2 Coefficients Least Squares Estimate 1.78707 1.38683 Parameter Intercept Slope Analysis of Variance Source Sum of Squares Model 5.41183 Residual 0.00737143 Total (Corr.) 5.4192 Standard Error 0.0314283 0.0228897 Df T Statistic 56.8618 60.5873 Mean Square 5.41183 0.00147429 P-Value 0.0000 0.0000 F-Ratio 3670.82 P-Value 0.0000 lu an n va ie gh tn to Correlation Coefficient = 0.99932 R-squared = 99.864 percent R-squared (adjusted for d.f.) = 99.8368 percent Standard Error of Est = 0.0383964 Mean absolute error = 0.0260195 Durbin-Watson statistic = 1.85824 (P=0.2187) Lag residual autocorrelation = -0.0171856 p The StatAdvisor The output shows the results of fitting a square root-Y logarithmic-X model to describe the relationship between H Do Vang and A The equation of the fitted model is oa nl w d H Do Vang = (1.78707 + 1.38683*ln(A))^2 lu u nf va an Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Do Vang and A at the 95.0% confidence level ll The R-Squared statistic indicates that the model as fitted explains 99.864% of the variability in H Do Vang after transforming to an inverse normal scale to linearize the model The correlation coefficient equals 0.99932, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.0383964 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu oi m z at nh z an Lu Simple Regression - H Do Vang vs A Dependent variable: H Do Vang (CCao Do vang) Independent variable: A (Tuoi) m co l gm @ The mean absolute error (MAE) of 0.0260195 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level n va ac th RR si Square root-X model: Y = a + b*sqrt(X) Coefficients Least Squares Estimate -6.94803 10.2864 Parameter Intercept Slope Analysis of Variance Source Sum of Squares Model 216.988 Residual 0.729279 Total (Corr.) 217.717 Standard Error 0.533383 0.266691 Df T Statistic -13.0263 38.5705 Mean Square 216.988 0.145856 P-Value 0.0000 0.0000 F-Ratio 1487.69 P-Value 0.0000 lu an n va ie gh tn to Correlation Coefficient = 0.998324 R-squared = 99.665 percent R-squared (adjusted for d.f.) = 99.598 percent Standard Error of Est = 0.381911 Mean absolute error = 0.263433 Durbin-Watson statistic = 1.29357 (P=0.0433) Lag residual autocorrelation = 0.190351 p The StatAdvisor The output shows the results of fitting a square root-X model to describe the relationship between H Do Vang and A The equation of the fitted model is oa nl w d H Do Vang = -6.94803 + 10.2864*sqrt(A) lu u nf va an Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Do Vang and A at the 95.0% confidence level ll The R-Squared statistic indicates that the model as fitted explains 99.665% of the variability in H Do Vang The correlation coefficient equals 0.998324, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.381911 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu oi m z at nh z an Lu Simple Regression - H Do Vang vs A Dependent variable: H Do Vang (CCao Do vang) Independent variable: A (Tuoi) m co l gm @ The mean absolute error (MAE) of 0.263433 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is less than 0.05, there is an indication of possible serial correlation at the 95.0% confidence level Plot the residuals versus row order to see if there is any pattern that can be seen n va ac th SS si Squared-Y model: Y = sqrt(a + b*X) Coefficients Least Squares Estimate -67.8557 66.0661 Parameter Intercept Slope Analysis of Variance Source Sum of Squares Model 122212 Residual 543.74 Total (Corr.) 122756 Standard Error 8.81346 1.97075 Df T Statistic -7.6991 33.5233 P-Value 0.0006 0.0000 Mean Square 122212 108.748 F-Ratio 1123.81 P-Value 0.0000 lu an n va ie gh tn to Correlation Coefficient = 0.997783 R-squared = 99.5571 percent R-squared (adjusted for d.f.) = 99.4685 percent Standard Error of Est = 10.4282 Mean absolute error = 7.55592 Durbin-Watson statistic = 1.75022 (P=0.1635) Lag residual autocorrelation = -0.014239 p The StatAdvisor The output shows the results of fitting a squared-Y model to describe the relationship between H Do Vang and A The equation of the fitted model is oa nl w d H Do Vang = sqrt(-67.8557 + 66.0661*A) lu u nf va an Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Do Vang and A at the 95.0% confidence level ll The R-Squared statistic indicates that the model as fitted explains 99.5571% of the variability in H Do Vang after transforming to a reciprocal scale to linearize the model The correlation coefficient equals 0.997783, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 10.4282 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu oi m z at nh z an Lu Simple Regression - H Do Vang vs A Dependent variable: H Do Vang (CCao Do vang) Independent variable: A (Tuoi) m co l gm @ The mean absolute error (MAE) of 7.55592 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level n va ac th TT si Double reciprocal model: Y = 1/(a + b/X) Coefficients Least Squares Estimate -0.00265686 0.30651 Parameter Intercept Slope Analysis of Variance Source Sum of Squares Model 0.0518015 Residual 0.000557884 Total (Corr.) 0.0523594 Standard Error 0.00661087 0.0142253 Df T Statistic -0.401893 21.5469 Mean Square 0.0518015 0.000111577 P-Value 0.7044 0.0000 F-Ratio 464.27 P-Value 0.0000 lu an n va ie gh tn to Correlation Coefficient = 0.994658 R-squared = 98.9345 percent R-squared (adjusted for d.f.) = 98.7214 percent Standard Error of Est = 0.010563 Mean absolute error = 0.00767671 Durbin-Watson statistic = 1.54799 (P=0.1235) Lag residual autocorrelation = 0.0802324 p The StatAdvisor The output shows the results of fitting a double reciprocal model to describe the relationship between H Do Vang and A The equation of the fitted model is oa nl w d H Do Vang = 1/(-0.00265686 + 0.30651/A) lu u nf va an Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Do Vang and A at the 95.0% confidence level ll The R-Squared statistic indicates that the model as fitted explains 98.9345% of the variability in H Do Vang The correlation coefficient equals 0.994658, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.010563 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu oi m z at nh z an Lu Simple Regression - H Do Vang vs A Dependent variable: H Do Vang (CCao Do vang) Independent variable: A (Tuoi) m co l gm @ The mean absolute error (MAE) of 0.00767671 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level n va ac th UU si Logarithmic-X model: Y = a + b*ln(X) Coefficients Least Squares Estimate 2.22554 8.72959 Parameter Intercept Slope Analysis of Variance Source Sum of Squares Model 214.431 Residual 3.28567 Total (Corr.) 217.717 Standard Error 0.663524 0.483256 Df T Statistic 3.35413 18.0641 Mean Square 214.431 0.657135 P-Value 0.0202 0.0000 F-Ratio 326.31 P-Value 0.0000 lu an n va ie gh tn to Correlation Coefficient = 0.992426 R-squared = 98.4909 percent R-squared (adjusted for d.f.) = 98.189 percent Standard Error of Est = 0.810639 Mean absolute error = 0.577071 Durbin-Watson statistic = 1.31784 (P=0.0506) Lag residual autocorrelation = 0.144064 p The StatAdvisor The output shows the results of fitting a logarithmic-X model to describe the relationship between H Do Vang and A The equation of the fitted model is oa nl w d H Do Vang = 2.22554 + 8.72959*ln(A) lu u nf va an Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Do Vang and A at the 95.0% confidence level ll The R-Squared statistic indicates that the model as fitted explains 98.4909% of the variability in H Do Vang after transforming to a Y/(1-Y) scale to linearize the model The correlation coefficient equals 0.992426, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.810639 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu oi m z at nh z m co l gm @ The mean absolute error (MAE) of 0.577071 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level an Lu n va ac th VV si Phụ lục KẾT QUẢ TƯƠNG QUAN GIỮA HVN VÀ A TRÊN NỀN ĐẤT ĐEN TỪ CÁC HÀM MẶC ĐỊNH TRONG STATGRAPHICS PLUS CENTURION XV.I Comparison of Alternative Models - H Den vs A lu an n va p ie gh tn to d oa nl w ll u nf va an lu R-Squared 99.78% 99.61% 99.49% 99.30% 98.71% 98.07% 98.00% 97.31% 97.19% 96.63% 95.69% 93.62% 92.90% 89.89% 86.47% 85.79% 84.16% 78.61% 68.84% 68.80% 67.39% 48.49% m Correlation 0.9989 0.9981 0.9974 0.9965 0.9935 0.9903 0.9900 0.9864 -0.9859 0.9830 0.9782 0.9675 0.9639 0.9481 0.9299 0.9262 -0.9174 0.8866 -0.8297 0.8295 -0.8209 -0.6964 oi z at nh z gm @ Model Square root-Y logarithmic-X Square root-X Double reciprocal Squared-Y Multiplicative Double square root Logarithmic-X Linear S-curve model Squared-Y square root-X Double squared Logarithmic-Y square root-X Square root-Y Squared-Y logarithmic-X Squared-X Exponential Reciprocal-X Square root-Y squared-X Squared-Y reciprocal-X Logarithmic-Y squared-X Reciprocal-Y Reciprocal-Y squared-X Reciprocal-Y square root-X Reciprocal-Y logarithmic-X Square root-Y reciprocal-X Logistic Log probit m co l The StatAdvisor This table shows the results of fitting several curvilinear models to the data Of the models fitted, the square root-Y logarithmic-X model yields the highest R-Squared value with 99.777% This is 2.47075% higher than the currently selected linear model To change models, select the Analysis Options dialog box an Lu n va ac th WW si Simple Regression - H Den vs A Dependent variable: H Den (CCao Den) Independent variable: A (Tuoi) Square root-Y logarithmic-X model: Y = (a + b*ln(X))^2 Coefficients Least Squares Estimate 1.81032 1.38327 Parameter Intercept Slope lu Analysis of Variance Source Sum of Squares Model 5.38409 Residual 0.0120309 Total (Corr.) 5.39612 Standard Error 0.0401507 0.0292424 an Df T Statistic 45.0881 47.3034 Mean Square 5.38409 0.00240618 P-Value 0.0000 0.0000 F-Ratio 2237.61 P-Value 0.0000 n va p ie gh tn to Correlation Coefficient = 0.998885 R-squared = 99.777 percent R-squared (adjusted for d.f.) = 99.7325 percent Standard Error of Est = 0.0490528 Mean absolute error = 0.0349685 Durbin-Watson statistic = 1.67482 (P=0.1445) Lag residual autocorrelation = 0.0785567 nl w d oa The StatAdvisor The output shows the results of fitting a square root-Y logarithmic-X model to describe the relationship between H Den and A The equation of the fitted model is an lu u nf va H Den = (1.81032 + 1.38327*ln(A))^2 ll Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Den and A at the 95.0% confidence level oi m z at nh The R-Squared statistic indicates that the model as fitted explains 99.777% of the variability in H Den after transforming to an inverse normal scale to linearize the model The correlation coefficient equals 0.998885, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.0490528 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu z gm @ m co l The mean absolute error (MAE) of 0.0349685 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level an Lu n va ac th XX si Simple Regression - H Den vs A Dependent variable: H Den (CCao Den) Independent variable: A (Tuoi) Square root-X model: Y = a + b*sqrt(X) Coefficients Least Squares Estimate -7.04106 10.4015 Parameter Intercept Slope lu Analysis of Variance Source Sum of Squares Model 221.871 Residual 0.857622 Total (Corr.) 222.729 Standard Error 0.578416 0.289208 an Df T Statistic -12.173 35.9656 P-Value 0.0001 0.0000 Mean Square 221.871 0.171524 F-Ratio 1293.52 P-Value 0.0000 n va p ie gh tn to Correlation Coefficient = 0.998073 R-squared = 99.6149 percent R-squared (adjusted for d.f.) = 99.5379 percent Standard Error of Est = 0.414155 Mean absolute error = 0.2743 Durbin-Watson statistic = 1.38198 (P=0.0590) Lag residual autocorrelation = 0.174437 nl w d oa The StatAdvisor The output shows the results of fitting a square root-X model to describe the relationship between H Den and A The equation of the fitted model is an lu u nf va H Den = -7.04106 + 10.4015*sqrt(A) ll Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Den and A at the 95.0% confidence level oi m z at nh The R-Squared statistic indicates that the model as fitted explains 99.6149% of the variability in H Den The correlation coefficient equals 0.998073, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.414155 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu z gm @ m co l The mean absolute error (MAE) of 0.2743 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level an Lu n va ac th YY si Simple Regression - H Den vs A Dependent variable: H Den (CCao Den) Independent variable: A (Tuoi) Double reciprocal model: Y = 1/(a + b/X) Coefficients Least Squares Estimate 0.00189169 0.286769 Parameter Intercept Slope lu Analysis of Variance Source Sum of Squares Model 0.0453437 Residual 0.000233897 Total (Corr.) 0.0455776 Standard Error 0.00428054 0.00921087 an Df T Statistic 0.441927 31.1338 Mean Square 0.0453437 0.0000467793 P-Value 0.6770 0.0000 F-Ratio 969.31 P-Value 0.0000 n va p ie gh tn to Correlation Coefficient = 0.997431 R-squared = 99.4868 percent R-squared (adjusted for d.f.) = 99.3842 percent Standard Error of Est = 0.00683954 Mean absolute error = 0.00493123 Durbin-Watson statistic = 1.33075 (P=0.0662) Lag residual autocorrelation = 0.161953 nl w d oa The StatAdvisor The output shows the results of fitting a double reciprocal model to describe the relationship between H Den and A The equation of the fitted model is an lu u nf va H Den = 1/(0.00189169 + 0.286769/A) ll Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Den and A at the 95.0% confidence level oi m z at nh The R-Squared statistic indicates that the model as fitted explains 99.4868% of the variability in H Den The correlation coefficient equals 0.997431, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.00683954 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu z gm @ m co l The mean absolute error (MAE) of 0.00493123 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level an Lu n va ac th ZZ si Simple Regression - H Den vs A Dependent variable: H Den (CCao Den) Independent variable: A (Tuoi) Squared-Y model: Y = sqrt(a + b*X) Coefficients Least Squares Estimate -71.2514 67.9246 Parameter Intercept Slope lu Analysis of Variance Source Sum of Squares Model 129185 Residual 910.528 Total (Corr.) 130096 Standard Error 11.4051 2.55025 an Df T Statistic -6.24735 26.6345 Mean Square 129185 182.106 P-Value 0.0015 0.0000 F-Ratio 709.40 P-Value 0.0000 n va p ie gh tn to Correlation Coefficient = 0.996494 R-squared = 99.3001 percent R-squared (adjusted for d.f.) = 99.1601 percent Standard Error of Est = 13.4947 Mean absolute error = 9.82531 Durbin-Watson statistic = 1.66404 (P=0.1322) Lag residual autocorrelation = 0.0364974 nl w d oa The StatAdvisor The output shows the results of fitting a squared-Y model to describe the relationship between H Den and A The equation of the fitted model is an lu u nf va H Den = sqrt(-71.2514 + 67.9246*A) ll Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Den and A at the 95.0% confidence level oi m z at nh The R-Squared statistic indicates that the model as fitted explains 99.3001% of the variability in H Den after transforming to a reciprocal scale to linearize the model The correlation coefficient equals 0.996494, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 13.4947 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu z gm @ m co l The mean absolute error (MAE) of 9.82531 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is greater than 0.05, there is no indication of serial autocorrelation in the residuals at the 95.0% confidence level an Lu n va ac th AAA si Simple Regression - H Den vs A Dependent variable: H Den (CCao Den) Independent variable: A (Tuoi) Multiplicative model: Y = a*X^b Coefficients Least Squares Estimate 1.30921 0.916671 Parameter Intercept Slope lu NOTE: intercept = ln(a) Analysis of Variance Source Sum of Squares Model 2.36444 Residual 0.0308578 Total (Corr.) 2.3953 Standard Error 0.0643024 0.0468325 an Mean Square 2.36444 0.00617156 P-Value 0.0000 0.0000 F-Ratio 383.12 P-Value 0.0000 n va Df T Statistic 20.3602 19.5734 p ie gh tn to Correlation Coefficient = 0.993538 R-squared = 98.7117 percent R-squared (adjusted for d.f.) = 98.4541 percent Standard Error of Est = 0.0785593 Mean absolute error = 0.0634264 Durbin-Watson statistic = 0.993983 (P=0.0131) Lag residual autocorrelation = 0.231521 oa nl w d The StatAdvisor The output shows the results of fitting a multiplicative model to describe the relationship between H Den and A The equation of the fitted model is va an lu ll or u nf H Den = exp(1.30921 + 0.916671*ln(A)) oi m z at nh ln(H Den) = 1.30921 + 0.916671*ln(A) Since the P-value in the ANOVA table is less than 0.05, there is a statistically significant relationship between H Den and A at the 95.0% confidence level z @ m co l gm The R-Squared statistic indicates that the model as fitted explains 98.7117% of the variability in H Den The correlation coefficient equals 0.993538, indicating a relatively strong relationship between the variables The standard error of the estimate shows the standard deviation of the residuals to be 0.0785593 This value can be used to construct prediction limits for new observations by selecting the Forecasts option from the text menu an Lu n va ac th BBB si The mean absolute error (MAE) of 0.0634264 is the average value of the residuals The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file Since the P-value is less than 0.05, there is an indication of possible serial correlation at the 95.0% confidence level Plot the residuals versus row order to see if there is any pattern that can be seen lu an n va p ie gh tn to d oa nl w ll u nf va an lu oi m z at nh z m co l gm @ an Lu n va ac th CCC si