Soil Moisture During Prolonged Wet and Dry Periods

From: Michigan State University Crop Advisory Team Alert Vol. 16, No. 15, August, 2001.

Topsoil Moisture Content During Wet and Dry Periods

Jeff Andresen, Geography

The 2001 growing season will go down in the records as one of extremes in terms of soil moisture variability. Following unusually wet and cool conditions early in the growing season (including all-time records for May precipitation totals at some locations), dry conditions developed during the latter half of June and continued through much of July. Climatologically, soil moisture content is a result of the balance of water inputs, largely precipitation, and losses, including evapotranspiration (the combined loss of water through plant transpiration and soil evaporation), runoff, and drainage into groundwater. For most (unirrigated) soils in Michigan, this balance is dominated by precipitation and evapotranspiration. On an annual basis, soil water content tends to reach a maximum level during the early spring (prior to warm temperatures and vegetative growth), gradually decreasing during the summer to a minimum level in the fall, followed by the over winter recharge period in which losses to evapotranspiration are low due to low temperatures and minimal vegetative growth.

Among the variables monitored at stations in the Michigan Automated Weather Network (MAWN) is volumetric soil moisture, which is expressed as the ratio of water to soil in the same volume. This is a soil-specific characteristic dependent of the textural make-up of the soil and the size and shapes of its pore spaces. It thus can vary widely from site to site, but typically ranges from about 0.10 to 0.35 in._water³/in._soil³ (or just in³/in³ for short). Volumetric soil moisture is estimated at MAWN stations by a 12-inch forked probe which measures the dielectric constant of the volume of soil surrounding the probe which in turn is strongly dependent on the amount of water content. The probes were installed vertically into grass-covered soil next to the station, thus estimating average soil moisture content in the top 12 inches of the profile (Note that beginning in late 2001, probes will be redeployed at each MAWN station to monitor soil moisture at two different depths, 4 inches and 10 inches).

Mean daily soil moisture contents in the top 12 inches of the soil profile and daily precipitation totals at the Albion MAWN station from the 1^st of April through the 16^th of July are displayed in Figure 1. The moisture content during this period briefly exceeded 0.30 in³/in³ following 4.0+ inches of rainfall over a 3-day period from the 14th-16th of May and fell to only 0.15 in³/in³ by the 16^th of July. The rapid rises and slower, logarithmic decreases of soil moisture are strongly associated with the frequency and magnitude of daily precipitation. A rule of thumb that can be used when considering moisture in a soil profile down to rooting depth is that the total amount of plant available water in the profile decreases by 10% of the amount left on the previous day (given no added rainfall or irrigation). Thus, there is a gradual reduction of the daily total water loss over time. This decreasing reduction produces a pattern similar to that observed in Figure 1. There are two major decreases in the study period: between late April and early May (a drop from 0.30 in³/in³ to 0.17 in³/in³), and between mid June and the end of the series (July 16^th), when the soil moisture drops from almost 0.28 in³/in³ to about 0.15 in³/in³.

Like many atmospheric variables and processes, there are more than one temporal cycle present over long periods of time. Volumetric soil moisture content is no exception. Besides an annual cycle associated with the changing seasons, a distinct diurnal cycle is also frequently present, as can be seen in the three-day series of hourly average volumetric moisture content values at the same site (Albion) for the 26th-28th of June, 2001 (Figure 2). During each of these three days, the water content steadily drops during the daylight hours followed by a partial recovery during evening hours. The daytime losses are due to evapotranspiration while the recovery is associated with capillary movement of water upwards from lower layers and with the presence of dew on the ground surface (the sensor monitors all water within an elliptical-shaped radius pattern around the probe forks). Weather conditions on these three days were very similar in terms of air and soil temperature, relative humidity, and wind speed. In humid climates such as those usually experienced in Michigan, the rate of evapotranspiration is strongly dependent on the levels of incident solar radiation (wind speed and relative humidity play relatively larger roles in more arid climates). From data at the Albion MAWN station daily solar radiation totals of 518, 567, and 640 langleys/day were recorded for the 26th-28th, which are positively correlated with the approximate daytime water losses from this layer of 0.15, 0.20, and 0.25 in³/in³ respectively.

So what does all this mean in terms of water management?

In order to make some sense of these values of volumetric soil moisture, we must also consider the physical characteristics of the soil as well as plant canopy characteristics (which determine the rate of evapotranspiration losses). The canopy above the soil profile was kept mowed to an approximately 4-inch height. For the sake of simplicity, we shall conveniently assume that the roots of this grass are all located within the top 12 inches layer being monitored (likely not too far from reality). A soil survey of Calhoun County identifies the soil series in the vicinity of the station as Kalamazoo Loam, which has relatively similar fractions of sand, clay, and silt (USDA/NRCS, 1997). Plant-available water in the top 12 inches of the profile was measured in a number of representative Kalamazoo soil profiles by NRCS, resulting in a mean field capacity (roughly the maximum amount of water the soil can hold at equilibrium with gravity) of approximately 0.28 in³/in³ and a wilting point (the moisture level at which permanent plant wilt occurs) of 0.08 in³/in³, which is a typical representative range of soil moisture contents for this soil series and not too dissimilar from the range of values taken from the sensor at Albion (Figure 1).

For the first of the two figures, the total water in the top 12 inches at the maximum moisture content of 0.30 in³/in³ on April 19^th totals 3.6 inches (12 in._soil * 0.30 in._water³/in._soil). Note that this value is above the level of field capacity from the NRCS lab data and, given the magnitude, probably is close to the soil's saturation point in which all of the pores in the soil are filled with water. By the 11^th of May, the average daily value fell to 0.18 in³/in³ for a total decrease of 2.16 inches. Thus, the loss of water in the 12 inch layer over this relatively dry 21-day period is 3.6 inches - 2.16 inches = 1.44 inches, or an average of 0.07 inches per day. For the reasons mentioned above, this value is only an average and not representative of many of the days earlier in the drying period, when daily rates were greater. This is evident in Figure 2, where soil moisture levels were closer to the upper end of the range and daily decreases of 0.015, 0.020, and 0.025 in³/in³ were partially offset by additions of 0.005 in³/in³ for net 12-inch layer decreases of 0.12, 0.18, and 0.24 inches, respectively. Given no addition of any other water during this 3-day period (no runoff) and assuming minimal drainage losses , these values can then be thought of as good estimates of actual evapotranspiration from the canopy when it was relatively well-watered. At the end of the daily series in Figure 1, moisture levels off near 0.15 in³/in³ (just above the wilting point given by the NRCS lab data), likely because available water for the grass canopy has become very limited and the plants have partially shut down or become semi-dormant, in turn reducing the amount of water lost to the atmosphere in evapotranspiration.

In summary, the volumetric soil moisture probes offer some ability to monitor soil moisture content with more precision and at a lower cost than previous types of sensors. As with any technology, however, there are some potential drawbacks. First, the probes only take the top 12 inches of the soil profile into account. Many field crops grown in Michigan have deep rooting systems (e.g alfalfa, corn) potentially reaching down to a 6 foot depth or greater. In addition, soil moisture can vary widely over space, even within the same field. This is especially true in Michigan, where soil types vary from beach sand to heavy clay to organic muck. Thus, the numbers from the MAWN stations should serve as a general guide for topsoil layers only, with careful consideration of soil type. Additional probes may be installed at greater depths at MAWN stations in the future given sufficient resources.

Figure 1. Daily mean volumetric soil moisture (in³/in³) in the top 12 inches of the soil profile (under a 4-inch canopy) and precipitation (inches) from the MAWN automated weather station in Albion, MI, 1 April 2001 through 16 July, 2001.

Figure 2. Hourly mean volumetric soil moisture (in³/in³) in the top 12 inches of the soil profile (under a 4-inch grass canopy) from the MAWN automated weather station in Albion, MI, from 06:00, 26 June through 18:00, 28 June, 2001.