Looking at the Big Picture.
Existing hydrology and soil erosion models involve using characteristics such as bulk density, porosity, moisture content, hydraulic conductivity and infiltration rate.
These parameters are measured in the laboratory from soil core samples, in soil columns or in lysimeter tanks. Studies to determine soil erodibility and erosion rate are mainly conducted in experimental plots. The inherent soil variability under field conditions -- in space and time -- makes mathematical descriptions of soil structure and processes difficult and challenging.
Determining soil properties and hydraulic characteristics in core samples, soil columns and lysimeters and erosion parameters in experimental plots, are point measurement techniques. Studies at these scales reveal processes and relationships that may not be captured on a larger scale. However, soil variability in this micro-scale approach requires many measurements to obtain a representative value as the geographic area increases. It also requires extensive soil characterization.
Improved prediction accuracy can be achieved by measuring soil properties, hydraulic characteristics and erosion parameters on a larger scale. Macro-scale measurement can help reach this goal. The term refers to determining soil properties and hydraulic characteristics in an area or volume of soil larger than a core sample, soil column or lysimeter. It can also involve measuring erosion and runoff parameters in an area larger than an experimental plot.
The limited amount of soil samples that can be collected -- and high costs of laboratory analyses -- can preclude characterization of large tracts of land. When geographic area and depth increase, effects on soil parameters and processes occurring in soil are difficult to determine, especially in studies involving water flow and erosion.
Models and tools for spatial extrapolation, such as geographic information system (GIS) and satellite imagery, are improving research as computers become more powerful. However, input values are measured in minute sections of an interest area. The gap between measurement and application scales cannot be handled by point measurement techniques or by providing measures of variability and confidence limit.
Relative estimation values from modeling can help in comparing different farm management practices. This method provides farmers and policy makers with options for higher crop yield and economic return, better erosion control and improved water quality. However, more accurate prediction is needed for absolute values of peak runoff used to design dams, bridges, flood and erosion control structures, and of sediment and nutrient losses to set limits for total maximum daily loads (TMDLs) in streams.
Soil hydraulic characteristics and water flow are sensitive to boundary conditions at the perimeter of the area. As the size of measurement and observation area increases, the perimeter-to-area ratio decreases and reduces perimeter condition effects. This reduction provides data focusing on processes occurring within an area, rather than on its perimeter.
Water flow, soil detachment and soil deposition are affected by flow depth and velocity related to surface roughness, slope, length of slope and slope changes with distance. Rainfall and irrigation water cover large areas of land where water flows overland and through the soil profile regardless of soil characterization and agro-ecological zoning. Increasing the observation scale allows measurement to capture integrated effects of processes in the field and produces more representative values. These values match the hydrologic and erosion models' capability and intent for spatial extrapolation. Macro-scale measurement provides one observed value for a large area rather than several observations in smaller sub-sections of that area.
In a simulation exercise, coverage is divided into smaller, discrete areas or control volumes. A control volume is an area of soil surface and depth of soil profile. Its shape can follow available soil surface classification schemes, management zones, natural boundaries or any polygon. The simulation coverage can consist of control volumes with different shapes and sizes.
A control volume is studied in relation to adjacent control volumes considering interflow and interaction among them. To run a model, these subdivisions need all the values of required input parameters. Simulation models cover a wide range as long as the values of these inputs are available. The subdivision scale should not exceed the measurement and observation scale.
Regression analysis models are valid only within the range of independent variables. Similarly, applying parameter values measured in smaller areas is limited to the measurement area and can lead to errors when applied to larger areas. The size and shape of the control volume can define the measurement and observation scale. Sound experimentation principles including randomization, replication and appropriate design can then be used to determine parameter values and variability.
The simulation scale, specifically the spatial and temporal steps considered in the computation, should be within the scale of measurement. The time interval and soil depth in the experiment should be used in the simulation. Caution should be exercised in extrapolating experimental results. For example, water loss of 0.08 inch (2 millimeters) observed in two hours in an infiltrometer ring set-up can be converted to 0.94 inch (24 millimeters) per day; or 0.97 pounds (440 grams) of sediment loss from 2,368 square feet (220 square meters) of experimental plot can be converted to 17.8 pounds per acre (20 kilograms per hectare).
Although mathematically convenient, these extrapolations assume steady state conditions beyond actual time interval and observation area. A simple mathematical conversion can impose a new condition. Values of 0.04 inch (1 millimeter) per hour for water loss and 0.0004 pounds per square foot (2 grams per square meter) for soil loss are acceptable conversions because time and space intervals are less than actual. Extrapolation may be useful in data analysis and presentation of experimental results. But it may produce erroneous results when loosely applied in simulation studies.
Using models and computers, computations beyond actual time intervals and measurement space result in educated guesses and create impressive ambiance of confidence and reliability. But these advanced computational technologies fail to make up for the lack of field measurement techniques to obtain values more applicable to larger interest areas.
While soil scientists focus on core samples, lysimeters or plots, hydrologists are concerned about runoff from larger areas -- watersheds. A watershed, with defined boundary and outlet, is an observation unit where streamflow is measured and runoff losses due to erosion by rainfall and flowing water are estimated. The shape of the control volume is determined by the natural boundary.
Results from a 1997 watershed-scale erosion study in the Philippines, and four watersheds monitored in 1999 in northeast Iowa, showed correlation coefficients greater than 0.8 between rainfall and peak runoff, and peak runoff and total soil loss from runoff. These analyses demonstrate:
* parameter measurement at a scale larger than experimental plot and
* selection and use of macro-scale characteristics and responses. Macro-scale parameters had been defined in earlier hydrologic and groundwater flow models measured or estimated in the field.
The practical scale of measuring soil properties and characteristics at the macro level is a research question itself. It can be affected by methods, instrumentation and costs so observation and instrumentation techniques must be developed accordingly. The scale of parameter measurement dictates model development and scope of application. Approaches and techniques for collecting input values must cope with rapid advances in modeling and computer capabilities.
Macro-scale measurements and macro-scale parameters could direct hydrologic and soil erosion research through the 21st century.
ASAE member Marl us M. Agua is a post doctoral research associate in the agricultural and biosystems engineering department.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||soil research techniques|
|Author:||Agua, Marius M.|
|Publication:||Resource: Engineering & Technology for a Sustainable World|
|Date:||Sep 1, 2000|
|Previous Article:||A Chicken House Dust Buster.|
|Next Article:||Spectroscopic Analysis.|