Questioning the Basics
Managing any complex process involves observation, revision, and recalculation, and sometimes long-held assumptions get turned upside down. A recent study is raising questions about something that had seemed inviolate in our field: the Universal Soil Loss Equation, or USLE, and its successors.
Since the 1940s, researchers have been working on ways to estimate and quantify soil loss on agricultural lands. The original USLE was published in the 1960s and revised in 1978; the most recent version, RUSLE2, uses the same formula as USLE but includes some refinements to the determining factors. (The factors taken into account in both the USLE and RUSLE equations include rainfall-runoff erosivity, soil erodability, slope length, slope steepness, cover management, and erosion control practices). The equation, and the many models that have been derived from it, estimate average soil loss in tons per acre per year.
The recent study, conducted in the Chesapeake Bay watershed, examined how accurate USLE- and RUSLE2-derived predictions really are. One of the things the equations and related models have been used to predict is how much sediment will end up in rivers, lakes, and other surface waters, which is a critical piece of information for developing total maximum daily loads. Researchers compared actual sediment measurements in 78 catchments throughout the Chesapeake Bay watershed with USLE and RUSLE2 predictions, as well as predictions from five sediment-delivery algorithms, of the average annual sediment delivery to those catchments. They also did similar comparisons for another 23 catchments that are monitored by the US Geological Survey. Surprisingly, they found that the predictions of sediment yield exceeded the actual observed sediment yields by more than 100%.
The problem, they concluded, was that although RUSLE2 and the various models based on it might be good at predicting how much soil is lost to erosion, not all the eroded soil reaches streams, lakes, or in this case the Chesapeake Bay. Many things can affect how far the sediment travels and where it ends up, so the estimates of soil loss do not correlate well to sediment delivery rates. Although different models use various adjustments to compensate, they still aren’t very accurate at predicting how much sediment ends up in the water.
Advertisement
USLE and RUSLE have been such staples in the practice of erosion control that it’s unsettling to see them seemingly discredited like this. Actually, though, the researchers—Kathleen Boomer, Donald Weller, and Thomas Jordan of the Smithsonian Environmental Research Center—aren’t saying so much that USLE and its successors are wrong or that they don’t work for their original purpose, simply that they’re being misapplied—the wrong tools for the job. And that means there’s a great deal more work ahead to come up with more accurate prediction methods.
The study was published in the January/February 2008 issue of the Journal of Environmental Quality (http://jeq.scijournals.org).
Author's Bio: Janice Kaspersen is the editor of Erosion Control magazine.
May 2008
Questioning the Basics
Managing any complex process involves observation, revision, and recalculation, and sometimes long-held assumptions get turned upside down. A recent study is raising questions about something that had seemed inviolate in our field: the Universal Soil Loss Equation, or USLE, and its successors.
Since the 1940s, researchers have been working on ways to estimate and quantify soil loss on agricultural lands. The original USLE was published in the 1960s and revised in 1978; the most recent version, RUSLE2, uses the same formula as USLE but includes some refinements to the determining factors. (The factors taken into account in both the USLE and RUSLE equations include rainfall-runoff erosivity, soil erodability, slope length, slope steepness, cover management, and erosion control practices). The equation, and the many models that have been derived from it, estimate average soil loss in tons per acre per year.
The recent study, conducted in the Chesapeake Bay watershed, examined how accurate USLE- and RUSLE2-derived predictions really are. One of the things the equations and related models have been used to predict is how much sediment will end up in rivers, lakes, and other surface waters, which is a critical piece of information for developing total maximum daily loads. Researchers compared actual sediment measurements in 78 catchments throughout the Chesapeake Bay watershed with USLE and RUSLE2 predictions, as well as predictions from five sediment-delivery algorithms, of the average annual sediment delivery to those catchments. They also did similar comparisons for another 23 catchments that are monitored by the US Geological Survey. Surprisingly, they found that the predictions of sediment yield exceeded the actual observed sediment yields by more than 100%.
The problem, they concluded, was that although RUSLE2 and the various models based on it might be good at predicting how much soil is lost to erosion, not all the eroded soil reaches streams, lakes, or in this case the Chesapeake Bay. Many things can affect how far the sediment travels and where it ends up, so the estimates of soil loss do not correlate well to sediment delivery rates. Although different models use various adjustments to compensate, they still aren’t very accurate at predicting how much sediment ends up in the water.
USLE and RUSLE have been such staples in the practice of erosion control that it’s unsettling to see them seemingly discredited like this. Actually, though, the researchers—Kathleen Boomer, Donald Weller, and Thomas Jordan of the Smithsonian Environmental Research Center—aren’t saying so much that USLE and its successors are wrong or that they don’t work for their original purpose, simply that they’re being misapplied—the wrong tools for the job. And that means there’s a great deal more work ahead to come up with more accurate prediction methods.
The study was published in the January/February 2008 issue of the Journal of Environmental Quality (http://jeq.scijournals.org).