Why Numerical Intersections Need More Than a Graph
When two curves intersect, their \(x\)-coordinates are roots of the difference of their functions. For simple equations, those roots may be found exactly. In many applications, however, the equation cannot be solved conveniently by hand. A graphing calculator can estimate the intersections, but a graph alone is not enough: its viewing window and pixel resolution can hide a crossing, make a tangency look like a crossing, or suggest more precision than the estimate supports.
For an area calculation, the goal is not merely to find a list of approximate intersections. You must identify which intersections bound the particular region, establish a useful interval for each numerical root, and make sure the area you report is consistent with the accuracy of those bounds.
A Reliable Calculator Workflow
Define the difference function \(h(x)=f(x)-g(x)\). The curves intersect wherever \(h(x)=0\). On a graphing calculator, you can either graph both original functions and use an intersection feature, or graph \(h\) and use a zero feature. In either case, record the domain being investigated and inspect a table of values near each candidate.
A candidate root is more convincing when values of \(h\) on either side are small and have opposite signs. If \(h\) is continuous, opposite signs at the endpoints of an interval guarantee at least one root between them by the Intermediate Value Theorem. They do not, by themselves, guarantee that there is only one root in that interval, so zooming in or checking the graph and nearby values remains useful. A tangency may not change sign at all; therefore, a sign-change test alone can miss an intersection.
Use the interval relevant to the region. A wide window may help reveal the overall shape; a narrower window helps inspect individual intersections.
Graph both curves or graph their difference. Use a calculator’s intersection or zero feature to obtain approximate locations.
Evaluate \(h(x)\) at nearby values. Record a bracket whose endpoints lie on opposite sides of a sign-changing root, or use an appropriately refined numerical estimate for a tangency.
Use the stated domain and the graph’s crossing order. Do not automatically use every root visible on the calculator.
Use the selected bounds in the area integral. Report only digits supported by the root estimates and numerical evaluation.
Worked Example: A Line and an Exponential Curve
Worked Example: Isolating the Positive Intersection
Find a numerical bound for the area enclosed by \(y=2x+1\) and \(y=e^x\) between their intersections at \(x=0\) and \(x=r\), where \(r>0\).
Set the functions equal, or equivalently define \(h(x)=e^x-(2x+1)\). We know \(h(0)=1-1=0\), so one intersection is exactly \(x=0\). To locate the other, use the calculator on the positive \(x\)-axis. For example, \(h(1.25)\approx-0.009657\), while \(h(1.26)\approx0.005421\). Since \(h\) is continuous, a root lies between \(1.25\) and \(1.26\). Refining the calculator search gives \(r\approx1.2564312\).
On the interval from \(0\) to \(r\), the line is above the exponential curve: for example, at \(x=1\), \(2x+1=3\) and \(e^1\approx2.71828\). Since \(h'(x)=e^x-2\), \(h\) decreases until \(x=\ln 2\) and increases afterward; with \(h(0)=0\) and \(h(\ln 2)<0\), there is exactly one positive root \(r\), and \(h<0\) between \(0\) and \(r\). So the region bounded by the two intersections lies between \(x=0\) and \(x=r\). Its area is
An antiderivative is \(x^2+x-e^x\), so
At the intersection, \(e^r=2r+1\). Substituting this relation gives
Using the more precise root estimate, \(A\approx(1.2564312)^2-1.2564312\approx0.322188\). Thus the area is approximately \(0.3222\) square units, to four decimal places.
This estimate can also be bounded using a root bracket. Refinement gives \(1.2564<r<1.2565\). Since \(r^2-r\) is increasing for \(r>1/2\), the area lies between its values at the bracket endpoints:
The bracket supports a report of about \(0.322\) to three decimal places. The refined root estimate supports the more precise calculator estimate above.
Why Root Brackets Help Bound Area
A calculator’s root estimate is an approximation, not an exact endpoint. When the area can be written as a function of that endpoint, a root bracket can be carried through the calculation. For example, in the previous problem the area was \(A(r)=r^2-r\), and \(A'(r)=2r-1>0\) throughout the root bracket. Therefore, increasing \(r\) increases the area, so evaluating the area formula at the two ends of the bracket gives lower and upper bounds.
More generally, if an area is \(A(r)=\int_a^r q(x)\,dx\) and \(q(x)\geq0\) on the interval under consideration, then \(A\) increases as \(r\) increases. Indeed, if \(r_1<r_2\), then
Thus, if \(r_L<r<r_U\), then \(A(r_L)\leq A(r)\leq A(r_U)\), provided the same nonnegative integrand describes the area throughout those intervals. The condition matters: if the integrand changes sign, the area function need not increase. In that case, check its behavior rather than assuming endpoint evaluations give the bounds.
Worked Example: Selecting the Second Intersection
Worked Example: A Logarithm and a Line
Consider \(y=\ln(x+1)\) and \(y=x/2\) for \(x\geq0\). Estimate the area between the curves from their intersection at \(x=0\) to their next intersection.
Define \(h(x)=\ln(x+1)-x/2\). The first root is \(h(0)=0\). Calculator values show \(h(2.5128)>0\) and \(h(2.5130)<0\), isolating the next root in that narrow interval. A refined numerical estimate is \(r\approx2.51286\). The root at zero is not the upper bound: the requested region continues until the next intersection.
At \(x=1\), \(\ln 2\approx0.6931>0.5\), so the logarithmic curve is above the line between the two roots. The area is
An antiderivative of \(\ln(x+1)\) is \((x+1)\ln(x+1)-(x+1)\), and an antiderivative of \(x/2\) is \(x^2/4\). Therefore,
At the nonzero intersection, \(\ln(r+1)=r/2\). Substitution simplifies the expression:
Using \(r\approx2.51286\), this gives \(A\approx2.51286(0.51286)/4\approx0.32219\) square units. For a bracket-based check, \(r\) lies between \(2.5128\) and \(2.5130\). The function \(r(r-2)/4\) is increasing for \(r>1\), since its derivative is \((r-1)/2\). Thus
The numerical area is therefore approximately \(0.3222\) square units to four decimal places. The bracket confirms the estimate to three decimal places; the refined numerical calculation supplies the additional digit.
Several Intersections on a Specified Interval
The intended region determines which roots matter. A graph may show several intersections, but only the ones that form the boundary of the requested region belong in that area calculation. Also, a curve can touch another curve and turn around without changing which one is above. A graphing calculator’s intersection feature can find such a contact, while a sign-change search may fail to detect it. Inspect the graph and table near suspected tangencies as well as near crossings.
Worked Example: Two Relevant Crossings of a Horizontal Line
On \(0\leq x\leq\pi\), find the area between \(y=\sin x\) and \(y=0.2\). A calculator or inverse-sine evaluation gives the intersections \(a\approx0.2013579\) and \(b\approx2.9402347\). Both roots are needed: they bound the part of the interval where the sine curve lies above the horizontal line.
The area is
An antiderivative is \(-\cos x-0.2x\). Because \(\sin a=\sin b=0.2\), the calculator values are approximately \(\cos a=0.9797959\) and \(\cos b=-0.9797959\). Substituting the endpoints gives
To four decimal places, the area is \(1.4118\) square units. As a check, the derivative of the antiderivative \(-\cos x-0.2x\) is \(\sin x-0.2\), which agrees with the integrand.
Common Mistakes and Reporting Precision
A frequent mistake is to read one approximate intersection from the graph and treat it as an exact endpoint. Another is to choose the first root found by the calculator even when the region extends to a later crossing. These errors can change both the integral and the region represented by it.
Always keep enough digits in intermediate calculations and round at the end. A root estimate and an area estimate have different precision: small uncertainty in an endpoint may have little effect on an area, or it may affect the last reported digit. When the question asks for numerical bounds, give an interval or a rounded value whose accuracy is supported by your checks. State units when the axes represent quantities with units; otherwise, describe the result as square units.
Check Your Understanding
For each question, explain your reasoning, not only the calculator key or numerical result.
- What difference function would you graph to locate intersections of \(y=f(x)\) and \(y=g(x)\)?
- Why does a sign change in a continuous difference function guarantee a root, and why might a tangency be missed by that test?
- If the region ends at the second positive intersection, why is it unsafe to use the first root returned by a calculator’s zero feature?
- A root is bracketed by \(r_L<r<r_U\), and \(A(r)\) is increasing on that bracket. Which endpoint gives the lower area bound, and which gives the upper bound?
- Why should you retain extra digits in a root estimate and round the area only at the final step?
