THE UNIVERSITY OF LETHBRIDGE DEPARTMENT OF GEOGRAPHY GEOGRAPHY 3235: Quantitative Models for Geographic Analysis Fall 2006 Assignment 2 – Trend Projection (Part 2) Goal To use several types of trend projection procedures to fit an appropriate curve to historical data and to critically evaluate the output. Data The data for this exercise are provided to you in an Excel spreadsheet: University of Lethbridge Full-time Fall Students – Source: University Statistics and Fact Book publications, University of Lethbridge Archives. Yearly Average Carbon Dioxide Emissions – Source: C. D. Keeling, T.P. Whorf, and the Carbon Dioxide Research Group, Scripps Institution of Oceanography (SIO), University of California. Ridgewood Heights – Source: City of Lethbridge Planning Department, Municipal Census, various years. Using the Template To fit a linear curve, compute the coefficients using formulas, use the LINEST array function, run the regression tool (Tools>Data analysis>Regression) routine on index X and raw Y values, and use the trend line procedure after you have graphed the data (question 4) (Chart>Add trend line). Compute the projected values using the coefficients calculated above (they should all be the same!) and check your results using the TREND array function. To fit a geometric curve, compute the coefficients using formulas, use the LOGEST array function, run the regression data analysis routine on index X and log (Y) values and use the trend line procedure after you have graphed the data (question 4) (Chart>Add trend line). Compute the projected values using the coefficients calculated above and check your results using the GROWTH array function. Page 1 of 5 THE UNIVERSITY OF LETHBRIDGE DEPARTMENT OF GEOGRAPHY GEOGRAPHY 3235: Quantitative Models for Geographic Analysis Fall 2006 To fit a modified exponential curve, compute the coefficients using formulas and run the regression data analysis routine on index X and log (c-Y) values. (Note that it will return logs of the a and b coefficients.) Use the antilogs of these coefficients to compute projected values and check your work using the GROWTH array function using cfunction (c-Y). To fit a logistic curve, compute the coefficients using formulas and run the regression data analysis routine on index X and log (1/Y-1/c) values. Use the antilogs of these coefficients to compute projected values and check your work using the GROWTH array function using c-function (1/c-1/Y). Instructions 1. Using the Excel template provided, fit each of the four curves to the population data for the University of Lethbridge Full-time Fall Students data up to 2010. 2. Display and examine population projections, regression coefficients, and evaluation statistics. Create a table that shows the observed data and the projected data according to each curve, a summary measure for input evaluation, and a measure of goodness of fit. 3. Using a hand calculator, calculate YC values for 2010 and show your handwritten work– remember to use the index numbers not the actual years! a. Were you able to independently verify that the model is working properly? b. Why bother with this step? 4. Create a single X,Y graph that plots the observed points (without connecting lines) and the four curves (omitting the markers). Print your graph. 5. While you printed all four curves (just for practice), you should have selected only one. a. What factors led to your choice of curve? b. Discuss the results of your projection. 6. It is now Fall 2006. We know that the University of Lethbridge fall enrolment figures show a total enrolment of 8,000 students. a. With the clarity of hindsight, was your projection entirely satisfactory? b. How can you explain the differences? Page 2 of 5 THE UNIVERSITY OF LETHBRIDGE DEPARTMENT OF GEOGRAPHY GEOGRAPHY 3235: Quantitative Models for Geographic Analysis Fall 2006 7. Using data for Ridgewood Heights displayed in Table 3 in Part 1, imagine that the year is 2000 thus you only have data for 1984-1999 (***For a moment, ignore all population values after 1999***). Project Ridgewood’s population forwards to 2006. Use the Excel regression procedure to determine the a and b parameters and a spreadsheet to tabulate and graph your results. Remember, begun in 1984, the subdivision was planned for a total of approximately 530 dwelling units. 8. Think about the sort of process that underlies population growth in Ridgewood Heights. a. Is it really a demographic process up to 1999? b. Is this exercise pointless in some respects? Why or why not? (It may help to take a walk, drive or bike ride through Ridgewood Heights. Look for empty lots and space for future development.) 9. The municipal census of 2002 indicated that Ridgewood’s population dropped to 1,741 while its dwelling count increased to 536. By 2006 the population declined further to 1,615 while the number of dwellings increased to 538. Is the observed population consistent with your projection? Why or why not? As a forecaster, what additional factors might you have taken into account to adjust your projection when you were doing this work, way back in 1999? 10. Using the same procedure, project yearly average CO2 emissions at the Mauna Loa Observatory in Hawaii forward to the year 2025. Using your measures of input evaluation and goodness of fit, which curve is shown to be most appropriate in modelling this phenomenon? Present the projected yearly average CO2 emissions and the summary statistics for the most likely curve in a summary table. 11. Conclude by reflecting on the difference between a forecast and a projection. Based on your own judgement, would you discount any of the projections that you have produced? Why? How might you allow for these factors to arrive at a more probable forecast? Your laboratory report should be typed with a cover sheet and submitted to your lab instructor on or before October 5, 2006. Reports should be submitted only in person OR through the geography assignment drop box; no email submissions please. Page 3 of 5 THE UNIVERSITY OF LETHBRIDGE DEPARTMENT OF GEOGRAPHY GEOGRAPHY 3235: Quantitative Models for Geographic Analysis Fall 2006 Table 1 University of Lethbridge Full-time Fall Semester Students, 1967 - 2004 Year 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Full-Time Students 638 1024 1261 1409 1218 1076 1086 1154 1340 1474 1531 1439 1419 Year 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Full-Time Students 1502 1771 2208 2442 2633 2692 2763 2717 2937 3165 3548 3659 3776 Year 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Full-Time Students 3710 4247 4250 4447 4611 4725 5187 5544 5945 6175 6512 6847 Table 2 Yearly Average Carbon Dioxide Emissions, 1958 - 2004 Mauna Loa, Hawaii Year 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 CO2 Concentration (ppmv) 315.33 315.98 316.91 317.65 318.46 318.99 319.15 320.03 321.37 322.18 323.05 324.62 325.68 326.32 327.46 329.68 Year 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 CO2 Concentration (ppmv) 330.25 331.15 332.15 333.9 335.5 336.85 338.68 339.93 341.13 342.78 344.42 345.91 347.15 348.93 351.48 352.91 Page 4 of 5 Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 CO2 Concentration (ppmv) 354.19 355.59 356.37 357.04 358.89 360.88 362.64 363.76 366.63 368.31 369.48 371.02 373.1 375.64 377.38 THE UNIVERSITY OF LETHBRIDGE DEPARTMENT OF GEOGRAPHY GEOGRAPHY 3235: Quantitative Models for Geographic Analysis Fall 2006 Table 3 Ridgewood Heights Subdivision, Lethbridge Population and Dwelling Counts, Selected Census Years Year 1984 1985 1986 1987 1989 1992 1994 1997 1999 2002 2005 2006 Population 6 216 347 544 1006 1,617 1717 1766 1802 1741 1649 1615 Page 5 of 5 Dwelling Units 2 65 112 185 312 477 499 519 533 536 537 538