A watershed model uses rainfall, slope, and last year's land-cover map to predict flooding. A new parking lot is built after the map was made. What limits the model's prediction most?

An air-quality model combines factory emissions, wind direction, and valley topography. What can its output most directly support?

A population model projects a town's population using constant birth, death, and migration rates. Which conclusion is justified?

A lake model predicts 30% less phosphorus entering a lake if fertilizer application is reduced. What does this model result support?

A species-distribution model uses projected temperature and precipitation to map future suitable habitat. What claim can it support?

Climate models are run under several greenhouse-gas emissions pathways. What is an appropriate use of comparing their outputs?

A sea-level projection for one emissions pathway is 0.4 to 0.8 meters by 2100. What does this range express?

A laboratory reports a nitrate concentration of 12 ± 2 mg/L. How should this result be interpreted?

A drought model estimates a 10% to 25% chance of water restrictions next summer. What does the wide range indicate?

Two models using the same emissions pathway project warming of 1.5 to 2.1 °C. What is one reason their projections can differ?

A monitoring report gives a 95% confidence interval of 4 to 6 mg/L for Site A and 5 to 7 mg/L for Site B. What conclusion is supported by these intervals alone?

Hazard A is very likely but causes small harm each time; Hazard B is rare but can cause severe harm. How should their risks be compared?

Chemical A has an LD50 of 50 mg/kg, while Chemical B has an LD50 of 500 mg/kg. Chemical A is used at one-tenth the concentration of B. How should managers compare their risks?

Farm A has a 50% yearly chance of losing 1 metric ton of sediment. Farm B has a 5% yearly chance of losing 30 metric tons. Based on expected annual sediment loss, which farm has greater risk?

A pipeline rupture has a 2% annual chance and a $5 million cleanup cost. A recurring leak has a 40% annual chance and a $100,000 cleanup cost. Which has the larger expected annual cleanup cost?

A safety model estimates a 0.01 annual probability that a containment failure would kill 1,000 fish. What is the model's expected annual fish loss from this hazard?

A report says its flood-risk projection was peer-reviewed and includes its data, methods, and assumptions. A company summary gives a projection but shows no methods or data. Why is the first report more reliable?

A watershed model projects nitrogen runoff for 2035. What evidence would most directly strengthen confidence that the model is reliable?

A website claims a river is safe for the entire year based on two water samples collected in July. What is the main reliability problem with this claim?

Two groups publish wildfire-risk projections. One releases its raw data and analysis code, and the other releases only its conclusion. What makes the first projection more trustworthy?