Journal of Environmental Quality 31:1174-1183 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
TECHNICAL REPORTS
Landscape and Watershed Processes
Influence of Manure Application on Surface Energy and Snow Cover
Model Development and Sensitivities
C.E. Kongoli*,a and
W.L. Blandb
a NOAA/NESDIS/ORA, Atmospheric Research and Applications Division, 5200 Auth Rd., Rm. 601/WWB, Camp Springs, MD 20746-4304
b Dep. of Soil Science, 1525 Observatory Drive, Univ. of Wisconsin-Madison, Madison, WI 53706-1299
* Corresponding author (Cezar.Kongoli{at}noaa.gov)
Received for publication October 23, 2000.
 |
ABSTRACT
|
|---|
Winter landspreading is an important part of manure management in the U.S. Upper Midwest. Although the practice is thought to lead to excessive P runoff losses, surprisingly little has been learned from field experiments or current water quality models. We captured knowledge gained from winter manure landspreading experiments by modifying a mechanistic snow ablation model to include manure. The physically based, modified model simulated the observed delay in snow cover disappearance and surface energy balance changes caused by application of the manure. Additional model simulations of surface energy balance estimates of radiation and turbulent fluxes showed that during intense melting events the manure on top of snow significantly reduced the energy available for melt of the snow underneath, slowing melt. The effect was most pronounced when snowmelt was driven by both relatively high solar radiation and turbulent heat fluxes. High absorbed shortwave radiation caused significant warming of the manure, which led to substantial losses in turbulent fluxes and longwave radiation. Simulations of snowmelt also showed that manure applications between 45 and 100 Mg ha-1 significantly reduced peak snowmelt rates, in proportion to the manure applied. Lower snowmelt rates beneath manure may allow more infiltration of meltwater compared with bare snow. This infiltration and attenuated snowmelt runoff may partially mitigate the enhanced likelihood of P runoff from unincorporated winter-spread manure.
|
INTRODUCTION
|
|---|
RECYCLING OF LIVESTOCK manure in an environmentally sound way is a major challenge facing farmers in the northern USA. Even with generous public sector cost sharing, adoption of manure storage facilities appears to be possible for only a portion of the producers. As a result, landspreading in winter will remain an important part of manure management in this region for the foreseeable future. Although manure landspreading is thought to lead to excessive P runoff losses to surface water bodies, surprisingly little is known about how management options might reduce these losses. One such option is optimizing manure spreading through judicious decisions about timing, rate of application, and position on the landscape. This option involves no capital expense, and little, if any, additional labor.
Review of the literature on winter landspreading reveals that field experiments are extremely difficult, and that even very ambitious ones yield limited data. Compilations of such experiments may be found in Khaleel et al. (1980) and Moore and Madison (1985). Given the complexity of the problem, computer modeling may offer important opportunities to gain insight. Manure management options to minimize nutrient losses have usually been explored through the so-called hydrologic water quality (H/WQ) models such as WINHUSLE (Baun, 1995), CREAMS (Knisel, 1980), GLEAMS (Leonard et al., 1987, 1989), AGNPS (Young et al., 1987a,b; Tim and Jolly, 1994), EPIC (Sugiharto et al., 1994), ANSWERS (De Roo et al., 1989), and SWRRB-WQ (Arnold et al., 1990). These models have emphasized soil erosion and nutrient losses at the field and catchment scale. More specific models such as those by Wang et al. (1996) and Khaleel et al. (1979a)(b) target the transport of land-applied animal manure constituents. Moore and Madison (1985) developed a phosphorous loading model for winter-applied manure. Despite differences in complexity, common to all these models are a hydrology component to compute infiltration, drainage, and runoff, an erosion component to compute sediment transport, and a chemistry component to compute nutrient release and transport. Nutrient transport occurs by movement of sediment and/or water.
The hydrologic water quality models described above do not adequately address factors important to understanding the problem of winter manure management, such as manure effects on snow ablation, which have implications for P loss. In a companion paper we showed that solid manure applied on top of snow significantly delayed snow cover disappearance, in proportion to the manure applied (Kongoli and Bland, 2002). This delay was caused by changes to the surface energy balance. Current water quality models are not capable of representing the atmosphere–manure–snow system primarily because they lack a land–atmosphere exchange component and physically based representation of snow. Snow is merely regarded as storage for precipitation that can become infiltration or runoff (e.g., CREAMS, GLEAMS). Similarly, snowmelt is treated in a highly empirical fashion.
The problem of winter manure management should be studied in a mechanistic manner, and its complexity must be reduced to researchable components and then reintegrated into a model system. Steenhuis et al. (1981) provided an example of how one might start on a more focused and detailed model for the fate of N in winter-applied manure. They developed a field-scale model of the mechanics of nutrient release from manure applied on snow, and validated the model with laboratory and field experiments. The study ignored the micrometeorology of snowmelt as affected by the landspreading of manure, however. The only previous work to emphasize the significance of the micrometeorology of snowmelt as affected by manure was conducted by Braun (1990). His winter landspreading experiments in Switzerland showed that liquid manure applied on top of snow accelerated snowmelt, which exacerbated P loss in snowmelt runoff. He then modified an energy balance snow model to incorporate liquid manure, and used the model to explore implications of weather and application rates on snowmelt. The snowmelt model included energy exchange processes of radiation and turbulent fluxes, latent heat of fusion and thermal conduction in snow, and liquid water transport through the snowpack. The main hypothesis was that depending on the condition of the snow cover and the weather sequence, the risk of P loss in snowmelt runoff could be predicted, and the spreading of liquid manure avoided.
Braun's (1990) manure–snow model considered only the reduction of albedo by manure, which increased melt rates by increasing the net shortwave radiation flux to the snowpack. Our field experiments showed that the manure retarded melt (Kongoli and Bland, 2002). Drake (1980) developed a simplified, energy balance conceptual model of snowmelt as affected by surface dust to explain melt retardation. His work showed that changes caused to the surface albedo and temperature were both important to predicting whether snowmelt is retarded or advanced.
The main objective of the work reported here was to capture knowledge gained from manure landspreading experiments (Kongoli and Bland, 2002) in a mechanistic, numerical snow ablation model. This model would then allow exploration of implications of climate, timing, and manure application rates on snowmelt, an impossibility through field experiments or existing water quality models. In this study, we describe in detail the addition of manure to the snow ablation model. The snow model development and the atmosphere–land exchange routines are described in greater detail in Kongoli and Bland (2000) and Anderson et al. (1997)(2000). Here, we focus on the manure representation to simulate surface energy fluxes and snow cover as affected by the manure applied on top of snow. First, we describe in detail the conceptual framework and the assumptions made to represent the manure layer in the snow ablation model. Next, modeled snow depth, surface energy fluxes, and temperature estimates are compared with field measurements. Finally, we used the model to obtain additional estimates of surface energy balance and snow cover to better understand the winter manure system and the factors influencing the effect of manure on snowmelt.
This study is unique in that it is the first attempt to model winter manure systems in a mechanistic and comprehensive fashion. As mentioned earlier, these systems so far have been studied only empirically. The micrometeorology of snow is represented through a more sophisticated and extensively tested snow model than in the hydrologic water quality models. In addition, as will be described later, the manure–snow model developed here allows dynamic simulation of the atmosphere–manure–snow system in a variety of weather and snow cover conditions. Representation of manure P transport through the manure–snow–soil system in a mechanistic fashion requires further field experiments and modeling work. This new challenge is left for future research.
|
MATERIALS AND METHODS
|
|---|
The Snow Model
Earlier we developed and extensively validated a detailed mechanistic, numerical atmosphere–snow–soil model against 61 site years of continuous weather observations collected at three Wisconsin sites, and one site in Minnesota (Kongoli and Bland, 2000). The snow routine is based on Anderson's (1976) parameterizations and is numerically efficient and structurally robust, representing the snow and soil as a series of layers, defined by their unique physical and thermal properties. The snow routine represents the processes of snow accumulation, compaction, and metamorphosis, change of albedo, melt and liquid water retention and transport in snow, and energy balance. Results showed that with a minimum of calibration the model gave very good predictions of continuous snow depth, capturing critical processes of accumulation, ablation, and melt in a wide variety of situations.
Energy balance and heat transport routines were adopted from the Atmosphere–Land Exchange (ALEX) model of Anderson et al. (1997)( 2000). Energy exchange processes are represented in a physically based fashion, including radiation, sensible heat, and evaporation–sublimation between the soil–snow and the overlying air, and conduction through the snow and the underlying soil. The radiation balance is given by:
 | [1] |
where Rn is net radiation, Rin is incoming solar radiation at the surface, 
sf is surface albedo, Rin(1 - sf) represents net solar radiation, Lin is incoming atmospheric longwave radiation, and Lout is the outgoing longwave radiation emitted from the surface. All fluxes in Eq. [1] have units of W m-2.
Atmospheric longwave radiation incident on the surface, Lin, is given by:
 | [2] |
where 
ac is the atmospheric emissivity, 
is the Sephan–Bolzman constant (5.6697 x 10-8 W m-2 K-4), and Tk is air temperature (K). Outgoing longwave radiation, Lout, is given by:
 | [3] |
where sf is the surface emissivity and Tsf is surface temperature (K). The turbulent sensible heat flux is given by:
 | [4] |
Latent heat is given by:
 | [5] |
where subscripts a and sf signify properties of the air above and the surface material (soil or snow), respectively, T = temperature (K), e = vapor pressure (kPa), r = transport resistance (m s-1), 
= density of air (kg m-3), cp = heat capacity of air at constant pressure (J kg K-1), and 
= psychometric constant (kPa K-1). Fluxes in Eq. [4] and [5] have units of W m-2.
Transport of heat in the soil–snow profile occurs by conduction and is given by the time-dependent differential equation below (Campbell, 1985):
 | [6] |
where t = time (s), z = depth from the surface (m), c = volumetric heat capacity (J m-3 K-1), 
= thermal conductivity (J m-3 s-1 K-1), QH = heat source (W m-3), and 
z = layer thickness (m).
Incorporation of Manure
The snow model described above was modified to incorporate a single manure layer. Similar to the soil layers, the manure layer was represented as a distinct, solid, and stable material during the simulation period. This assumption is realistic only for solid or semi-solid manure (i.e., the manure that will not flow with perceptible movement without mechanical assistance) (Sobel, 1966). In contrast, semi-liquid or liquid manure may undergo drainage through the snow, a phenomenon modeled here only for meltwater. Although semi-solid and solid manures do not undergo movement by gravity forces, they may be subject to the processes of decomposition, or to the movement as a result of, for instance, heavy rainfall-runoff events or very strong winds. When manure is applied in winter, however, the processes of decomposition are slow (Khaleel et al., 1980). Also, the climatology of the Upper Midwest is such that heavy rainfall runoff is not likely to occur during winter conditions. For instance, during the field research on which this study is based (Kongoli and Bland, 2002), the amount of rainfall that fell during the winter was larger than the long-term average, but rainfall did not cause perceptible movement or changes to the structure of the manure. Also, we observed no significant change to the manure when new snow was deposited on top of it.
Although the applied manure was represented as a distinct material in the model, its location with respect to snow and soil surfaces undergoes continuous change as new snow is deposited on the manure, or the snow underneath disappears. When new snow is deposited on the manure, it undergoes processes of accumulation, densification, and melt similar to the snow beneath. The snow underneath the manure undergoes additional compaction due to the weight of the manure. Processes of compaction and metamorphosis are represented by the same equations as formulated by Anderson (1976) and demonstrated to be robust by Kongoli and Bland (2000). Meltwater resulting from snow on manure is allowed to drain through manure down to the layer underneath. Similarly, rainfall on manure is allowed to percolate down to the soil–snow underneath. Thus, changes in the water retained by manure are assumed negligible.
Similar to the snow and soil layers, heat transport in the manure is by conduction (Eq. [6]). Transmittance of radiation by the manure to the underlying material is assumed negligible. Although observations of radiative properties of manure are not available, measurements on residue materials suggest that transmission of radiation becomes negligible for bulk densities in the order of magnitude of those found in manure (e.g., Sauer, 1987). When computing surface fluxes of sensible and latent heat between manure and the overlying air (Eq. [4] and [5]), the surface roughness of the manure is set equal to that of a snow surface. Another assumption is that areas of manure-covered snow occur in small enough patches on the landscape such that characteristics of the bulk air (e.g., air temperature and vapor pressure) are not modified by the underlying surface changing from snow to manure.
Depending on application rate, manure coverage of the snow surface may be incomplete. When manure provides incomplete cover, it is assumed that an equivalent thickness of manure completely covers snow, but albedo and absorbed shortwave radiation (Eq. [1]) are modified to account for the uncovered portions within the manure. In this way, both mechanisms of heat transport through the manure and albedo reduction are represented in a similar fashion as areas completely covered by manure. Albedo (Eq. [1]) is computed as the area average of the manure and bare snow according to the expression:
 | [7] |
where mp is the albedo of the patchy area, sp is albedo of the bare snow, fsp is the fraction of bare snow, m is albedo of the manure, and fm is fraction of the manure. Outgoing longwave radiation is computed by Eq. [3], but Tsf is taken as the average of the bare snow and manure temperatures. We do not have a relationship between manure application rate and coverage except for what we observed in the experiments reported here. Our observations suggest that rates greater than 60 Mg ha-1 cover snow completely.
Weather and Site Inputs
Weather inputs to the model included hourly records of incoming solar radiation, air temperature, wet bulb temperature, wind speed, humidity, and daily values of total liquid precipitation, precipitation type, and cloud cover (Table 1)
. Precipitation type was derived from the daily snowfall obtained from the records of National Weather Service Cooperative Observers near the experimental sites. The fraction of cloud cover was assumed constant for the day, and was estimated from the measured solar radiation at the soil surface and the computed solar radiation incident on the outer edge of the atmosphere. Site inputs included initial snow water equivalent, elevation, latitude and longitude, slope and aspect, and soil type. The model estimates solar radiation on sloping surfaces from measured solar radiation by adjusting for slope, aspect, latitude, time of day, and time of year.
Manure Inputs
Manure inputs to the model included application rate, manure bulk density, manure coverage, albedo, heat capacity and conductivity, and time of application (Table 2)
. Based on application rate and bulk density, the model calculates the manure layer thickness. To simulate our 1998 and 1999 experiments, manure bulk density was derived from the average thickness of manure observed on completely covered plots and the application rate. Manure properties such as density, thickness, albedo, and thermal properties were held constant during the simulations. The manure thickness over the course of our experiments did not change significantly, so the assumption of a constant density was justified. Measured thermal conductivity of manure revealed fluctuations, but within relatively narrow ranges, so a mean value was selected as input and kept constant during simulations, but model sensitivity to this parameter was evaluated. For instance, all of the measured values on the 1998 and 1999 plots were between 0.08 and 0.2 W m-1 K-1. Highest values were found immediately after a rainfall event, but decreased in a few days and remained fairly constant (Kongoli and Bland, 2002). Heat capacity was estimated from Bohnhoff (1985) based on the type of manure and bulk density.
Model Verification and Computer Simulations
Verification of the modified manure–snow model was made against field measurements made in the winters of 1998 and 1999 at the Arlington Agricultural Research Station, Arlington, WI. Verification included surface temperature on the control and manured plots, surface net radiation and energy available for melt of the snow beneath the manure for representative days on manured plots, and daily snow depth and day of snow disappearance on control and manured plots. The model was also used to simulate additional surface energy balance components such as turbulent fluxes of sensible and latent heat on manured and control areas, which were not obtained from field experiments. Field measurements of turbulent fluxes of sensible and latent heat require open, extensive areas, so it is difficult to make these measurements for relatively small areas of manure. For all simulations, the model was initialized on 1 January (Day 1), before manure was applied. Manure was applied on Day 30 in 1998 at 70 Mg ha-1 and on Day 15 in 1999 at 45 (referred to as "patchy") and 100 ("heavy") Mg ha-1 (Table 2).
|
RESULTS AND DISCUSSION
|
|---|
We first compare observed and simulated snow depths and dates of disappearance. Then, a more detailed assessment is provided of the mechanistic operation of the model, by comparing observed and simulated surface temperature, energy available for melt of the snow beneath the manure, and radiation fluxes. Following this, we examine model simulations of fluxes that we did not measure. Finally, we use the model to explore effects of different spreading strategies.
Snow Depth and Snow Dynamics Simulation
The model captured manure effects on snow depth and snow cover disappearance well for both simulation years (Fig. 1 and 2)
. In 1998, the model accurately predicted complete snow cover disappearance in the manured plot on Day 59, but predicted snow disappearance in the control plots 3 d later than observed on Day 48 (Fig. 1). However, the model captured differences in snow dynamics between the control and manured plots remarkably well (e.g., until Day 44 there was little difference in measured snow depths). Snow dynamics was also generally well captured in 1999 on all the manured and control plots (Fig. 2), although in this year the manured plots were predicted to melt earlier than observed. For example, the snow disappeared on Day 35 on the control plot in the south-facing site, 4 d before the model predicted (Fig. 3A)
. Also, the snow in heavy plots disappeared on Day 42, but the model predicted this to occur completely 4 d later (Fig. 2).
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 2. Simulated and measured snow depths in the manured and control plots in 1999 for (A) the south-facing site and (B) the northwest-facing site.
|
|
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 3. Simulated and measured surface temperature in 1998 in (A) the manured plot and (B) the control plot. (C) Simulated surface temperature in the manured and control plots and recorded air temperature.
|
|
The model predicted no significant difference in snow depths and dynamics between the control plot in the south-facing site and the one in the northwest-facing site. We believe that the earlier melt observed for the south-facing site was because it accumulated surface impurities because of wind and dust, resulting in lower observed albedo than that of the north-facing site (Kongoli and Bland, 2002).
Surface Temperature and Radiant Flux Simulation
The model predicted reasonably well surface temperatures during the day on the control and manured plots during periods of rapid melting (i.e., Days 45 through 47 in 1998, and Days 32 through 35 in 1999) (Fig. 3A,B and 4A,B)
. Shown are days when manure was free of overlying snow. Melting of the bare snow is characterized by surface temperature of 0°C (e.g., Fig. 3B). The model captures the elevation of temperature by manure relative to the control (Fig. 3C and 4C). Accurate prediction of the difference in radiometric surface temperature between control and manured plots is a measure of the model's ability to simulate manure effects on the surface energy balance.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 4. Simulated and measured surface temperature in 1999 on the south-facing plots: (A) the heavy manured plot and (B) the control plot. (C) Simulated surface temperature in the manured and control plots and recorded air temperature.
|
|
Modeled net radiation (Fig. 5A and 6A)
and longwave radiation above the manure surface (Fig. 5B and 6B) and net energy conducted to the snowpack underneath the manure (Fig. 5C and 6C) agreed well with field measurements for representative days. Net energy to the snowpack underneath the manure was estimated as the difference between the conductive heat flux at the manure–snow interface and the soil–snow interface. As the soil–snow heat flux is negligible (Kongoli and Bland, 2002), net energy is essentially equal to the heat flux at the manure–snow interface. Net energy to the snowpack reflects the net radiation at the manure surface, but with reduced amplitude and slightly shifted phase. The unaccounted-for energy warms the manure and is lost back to the atmosphere by turbulent sensible and latent heat fluxes.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 5. Measured and simulated net (A) radiation, (B) longwave radiation, and (C) energy (conductive flux) to the snowpack beneath the manure in 1998.
|
|
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 6. Measured and simulated (A) net radiation, (B) net longwave radiation, and (C) net conductive flux to the snowpack beneath the heavy manure application, the south-facing plot in 1999.
|
|
Comparative Energy Balance of Snow and Manure–Snow Covers
To better understand melt dynamics as affected by manure, the validated model was used to produce complete energy balance estimates of snow and manure–snow covers. Shown are model estimates for representative days for simulations in 1998 and 1999 on a south-facing slope (Fig. 7–10
and Tables 3 and 4)
. As expected, the model showed consistently higher absorbed shortwave radiation by manure than by the snow surface (Fig. 7A and 9A). Modeled net radiation estimates are also higher in the manured plots than the control plots (Fig. 8A and 10A), although modeled net longwave radiation estimates are somewhat more negative, especially during the day (Fig. 7B and 9B). More negative values of longwave radiation during the day in the manured plots are because of the higher surface temperature caused by the manure. Higher daytime surface temperatures also led to significant losses in sensible and latent heat (turbulent fluxes) (Fig. 8B and 10B). Positive values of turbulent fluxes in these plots indicate energy loss from the manure, and negative values indicate gains. So, greater gains in net solar radiation by manure are associated with losses by turbulent fluxes, canceling out the albedo reduction effect of the manure. Furthermore, the net energy available for melt was generally smaller in manured plots than in the control plots (Fig. 8C and 10C). Net energy available for melt in the manured plots was significantly smaller than in the control plots in particular days (e.g., Days 46 and 47 in 1998 and Day 34 in 1999), and these were days of major melting. Thus, the manure acted as thermal insulator especially during intense melting, reducing the energy available for melt of the snow underneath. The insulating effect of manure caused the consistent melt retardation observed in the manured plots.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 7. Simulated net (A) shortwave radiation and (B) longwave radiation in the manured and control plots in 1998.
|
|
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 10. Simulated (A) net radiation, (B) turbulent flux (sum of sensible and latent heat), and (C) net energy of the snow beneath the manure in the south-facing heavy and control plots in 1999.
|
|
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 8. Simulated (A) net radiation, (B) turbulent flux (sum of sensible and latent heat), and (C) net energy of the snow beneath the manure in the manured and control plot in 1998.
|
|
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 9. Simulated (A) net shortwave radiation and (B) net longwave radiation in the south-facing heavy and control plots in 1999.
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Simulated net surface energy components (24-h sum) on the heavy-manured and control plots at the south-facing site in 1999.
|
|
The surface energy balance is key to understanding energy changes to the snow caused by manure (Tables 3 and 4). The tables give modeled estimates of daily net shortwave radiation, net longwave radiation, net turbulent fluxes of sensible and latent heat, and the residual of the surface energy transported by conduction for selected days. The energy residual S (W m-2 d-1) at the surface is computed from the equation:
 | [8] |
where Rsn is net shortwave radiation, Rln is net longwave radiation, and HS and HL are turbulent sensible and latent heat fluxes, respectively. Fluxes directed downward (toward the manure or snow surface) represent gains, and fluxes directed away from the surface represent losses. Under this convention, positive values of Rsn and Rln, or negative values of HS and HL, represent gains. Also, a positive value of S represents gain in surface energy. Days 46 and 47 in 1998 and Day 34 in 1999 show significantly higher values of the surface energy residual conducted to the control plots compared with manured areas. In Days 46 and 47 in 1998, the control plot experienced large gains from shortwave radiation and turbulent fluxes, and losses in longwave radiation (Table 3). Weather records indicated that these days had warm air temperatures (with a maximum of 5°C) and high incoming shortwave radiation. The warm air led to a strong supply of heat by turbulent fluxes to the control plots held at 0°C by melting, but not to the manured plots because their surface temperature was warmed by absorbed radiation. High shortwave radiation and turbulent fluxes caused rapid depletion of the snow in 1998 during these days. The manure slowed melting by causing the net surface energy on Days 46 and 47 to be much smaller (Table 3). Here, high gains in net shortwave radiation values caused by a lower manure albedo are associated with high losses in turbulent flux, whereas in the control plot the turbulent flux was a large gain. On Day 34 in 1999 (Table 4), all three components (i.e., shortwave radiation, longwave radiation and turbulent flux) contributed to the high gain in surface energy in the control plot (5.13 MJ m-2). In the manured plot, however, the gain in surface energy was significantly smaller, caused by the high loss in turbulent flux and no gain in longwave radiation.
Model Sensitivity to Spreading Strategies and Implications for Snowmelt
Sensitivity analysis of the manure–snow model involved simulations for the winters of 1998 and 1999 for a range of manure inputs (i.e., application rate, timing of application prior to and during the ablation phase, and thermal conductivity). The robustness of the manure–snow model with longer-term simulations was demonstrated in Kongoli (2000). Here, the weather dataset used as model input represented a wide variety of weather sequences and snow cover conditions. The robustness of the snow component was demonstrated through extensive sensitivity analysis and testing (Kongoli and Bland, 2000).
Timing of applications during ablation did not significantly affect simulations of snow depth and snowmelt. Shown are modeled snow depth and snowmelt estimates for three application rates in 1998 (Fig. 11)
and conductivities between 0.08 and 0.2 W m-1 K-1 observed in the field (Fig. 12)
. Starting time of manure applications was on Day 30 in 1998 and 15 in 1999, the same as in the field experiments (Kongoli and Bland, 2002). Modeled estimates of snow depth and snowmelt showed that at all rates of application manure significantly retarded melt (Fig. 11A) and reduced peak snowmelt rate on Day 47 (Fig. 11B). The majority of the reduction in peak melt rate was caused by the smallest manure application, 45 Mg ha-1 (about 60% coverage). Heavier manure applications further reduced the peak rate, but with less effectiveness than the lightest rate. A similar result was obtained from longer-term simulations at Madison, WI (Kongoli and Bland, 2000). This finding has implications for manure management. Light manure applications (45 Mg ha-1) reduce the quantity of manure P at risk for runoff compared with heavier applications, but are equally effective at reducing peak snowmelt rates.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 11. Simulated (A) snow depth and (B) rate of melt of the snow water equivalent depth underneath the manure in 1998 at different manure application rates. Simulations were made at a manure thermal conductivity (k) value of 0.14 W m-1 K-1.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 12. Simulated (A) snow depth and (B) rate of melt of the snow water equivalent depth underneath the manure in 1998 at different manure thermal conductivity (k) values (in units of W m-1 K-1). Simulations were made using at a manure application rate of 70 Mg ha-1.
|
|
Manure thermal conductivity had a significant effect on snow depth and the day of complete snow disappearance (Fig. 12), so it represents an important uncertainty in modeling the winter-manure system. However, conductivities within the observed range of 0.08 and 0.2 W m-1 K-1 retarded melt and significantly reduced the peak melt rate at a manure application rate of 70 Mg ha-1.
|
CONCLUSIONS
|
|---|
We modified a mechanistic numerical snow ablation model to simulate the surface energy balance and snow cover as affected by landspreading of manure. The original snow ablation model was extensively tested (Kongoli and Bland, 2000), and shown to adequately capture the micrometeorology of snow, essential to understanding the physics of winter manure systems.
The modified manure-on-snow model was verified with field measurements made on manure spread on top of snow during the winters of 1998 and 1999 in Arlington, WI. The model captured snow depths and dynamics generally well in both manured and control plots. Despite small discrepancies in simulating the day of complete snow disappearance, the model captured the observed melt retardation caused by the manure, and the underlying processes. Also, modeled estimates of surface temperature, fluxes of net radiation, and net fluxes of the snow pack underneath the manure agreed well with measured values. Modeled surface energy balance estimates of shortwave, longwave, and turbulent fluxes of sensible and latent heat revealed that during intense melting periods, manure significantly reduced the energy available to the underlying snow. This effect was most pronounced when snowmelt was driven by both radiation and turbulent fluxes. High solar radiation caused warming of the manure, which tended to minimize turbulent flux energy gain and increase loss in longwave radiation. Even relatively light applications of 45 Mg ha-1 gave substantial reductions in peak melt rate, which means springtime peak outflows may be decreased and delayed by manure applications.
|
ACKNOWLEDGMENTS
|
|---|
This research was supported by USDA-Hatch funds through project WIS03954, administered by the University of Wisconsin-Madison College of Agricultural and Life Sciences.
|
REFERENCES
|
|---|
- Anderson, E.A. 1976. A point energy and mass balance model of snow cover. NOAA Tech. Rep. NWS 19. U.S. Dep. of Commerce, Silver Spring, MD.
- Anderson, M.C., J.M. Norman, and G.R. Diak. 2000. ALEX—A reversible model of atmosphere–land exchange. Agric. For. Meteorol. 101:265–286.
- Anderson, M.C., J.M. Norman, G.R. Diak, W.P. Kustas, and J.R. Mecikalski. 1997. A two-source time-integrated model for estimating surface fluxes using thermal infrared remote sensing. Remote Sens. Environ. 60:195–216.
- Arnold, J.G., A.D. Williams, A.D. Nicks, and N.B. Sammons. 1990. SWRRB: A basin scale simulation model for soil and water resource management. Texas A&M Univ. Press, College Station.
- Baun, K. 1995. WINHUSLE for FOCS. Model documentation and user's manual. Version 2.0. Wisconsin Dep. of Natural Resour., Madison.
- Bohnhoff, D.R. 1985. Thermal and physical properties of separated manure properties. M.S. thesis. Univ. of Wisconsin, Madison.
- Braun, M. 1990. Zusammenhange zwischen Scheeedecke, gefrorenem Boden und Gullenabschemmung. (In German.) Diss. no. 9170. ETH Zurich.
- Bristow, K.L., G.S. Campbell, R.I. Papendick, and L.F. Elliot. 1986. Simulation of heat and moisture transfer through a surface residue–soil system. Agric. For. Meteor. 36:193–214.
- Campbell, G.S. 1985. Soil physics with BASIC: Transport models for soil–plant systems. Elsevier Science Publ. Co., Amsterdam.
- Campbell, G.S., and J.M. Norman. 1998. An introduction to environmental biophysics. Springer–Verlag, New York.
- De Roo, A.P.J., L. Hazelhoff, and P.A. Burrough. 1989. Soil erosion modeling using ANSWERS and geographic information systems. Earth Surf. Processes Landforms 14:517–532.
- Drake, J.J. 1980. The effects of surface dust on snowmelt rates. Arctic Alpine Res. 6:219–223.
- Flerchinger, G.N. 1995. SHAW—Simultaneous heat and water model—Technical documentation for SHAW. Version 2.1. Northwest Watershed Res. Center, Agric. Res. Serv., Boise, ID.
- Flerchinger, G.N. 1997. SHAW—Simultaneous heat and water model—Technical documentation for SHAW. Version 3.2. Northwest Watershed Res. Center, Agric. Res. Serv., Boise, ID.
- Khaleel, R., G.R. Foster, K.R. Reddy, M.R. Overcash, and P.W. Westerman. 1979a. A non-point source model for land areas receiving animal wastes: III. A conceptual model for sediment and manure transport. Trans. ASAE 22:1353–1361.
- Khaleel, R., G.R. Foster, K.R. Reddy, M.R. Overcash, and P.W. Westerman. 1979b. A non-point source model for land areas receiving animal wastes: IV. Model inputs and verification for sediment and manure transport. Trans. ASAE 22:1362–1368.
- Khaleel, R., K.R. Reddy, and M.R. Overcash. 1980. Transport of potential pollutants in runoff water from land areas receiving animal wastes. Water Res. 14:421–426.
- Knisel, W.G. 1980. CREAMS: A field-scale model for chemicals, runoff, and erosion from agricultural management systems. USDA Conserv. Res. Rep. 26. USDA-ARS, Washington, DC.
- Kongoli, C.E. 2000. Energy balance and ablation of snow as affected by manure application. Ph.D. thesis. Univ. of Wisconsin, Madison.
- Kongoli, C.E., and W.L. Bland. 2000. Modification of an atmosphere–land exchange model for simulating snow depths. Agric. For. Meteorol. 104:273–287.
- Kongoli, C.E., and W.L. Bland. 2002. Influence of manure application on surface energy balance and snow cover: Field experiments. J. Environ. Qual. 31:1166–1173 (this issue).[Abstract/Free Full Text]
- Leonard, R.A., W.G. Knisel, and D.A. Still. 1987. GLEAMS, Groundwater loading effects of agricultural management systems. Trans. ASAE 30:1403–1418.
- Leonard, R.A., H.F. Perkins, and W.G. Knisel. 1989. Relating agrochemical runoff and leaching to soil taxonomy: A GLEAMS model analysis. p. 158–160. In Proc. 1989 Georgia Water Res. Conf., Univ. of Georgia, Athens, GA. 16–17 May 1989. Univ. of Georgia, Athens.
- Monteith, J.L., and M.H. Unsworth. 1990. Principles of environmental physics. 2nd ed. E. Arnold, London.
- Moore, I.C., and F.W. Madison. 1985. Description and application of an animal waste phosphorous loading model. J. Environ. Qual. 14:364–369.[Abstract/Free Full Text]
- Sauer, T.J. 1993. Sensible and latent heat exchange at the soil surface beneath a maize canopy. Ph.D. thesis. Univ. of Wisconsin, Madison.
- Sobel, A.T. 1966. Physical properties of animal manures associated with handling. p. 140–143. In Proc. Natl. Sym. on Animal Waste Management. ASAE Publ. no. SP.0366. Am. Soc. Agric. Eng., St. Joseph, MI.
- Steenhuis, T.S., G.D. Bubenzer, J.C. Converse, and M.F. Walter. 1981. Winter-spread manure nitrogen loss. Trans. ASAE 436–449.
- Sugiharto, T., T.H. McIntosh, and R.C. Uhrig. 1994. Modeling alternatives to reduce dairy farm and watershed nonpoint source pollution. J. Environ. Qual. 23:18–24.
- Tim, U.S., and R. Jolly. 1994. Evaluating agricultural nonpoint-source pollution using integrated GIS and hydrologic/water quality model. J. Environ. Qual. 23:25–35.
- Wang, Y., D.R. Edwards, T.C. Daniel, and H.D. Scott. 1996. Simulation of runoff transport of animal manure constituents. Trans. ASAE 39:1367–1378.
- Young, R.A., C.A. Onstad, D.D. Bosch, and W.P. Anderson. 1987a. AGNSP: Agricultural non-point source pollution model. A watershed analysis tool. USDA Conserv. Res. Rep. 35. USDA-ARS, Washington, DC.
- Young, R.A., C.A. Onstad, D.D. Bosch, and W.P. Anderson. 1987b. AGNSP, a non-point source pollution model for evaluating agricultural watersheds. J. Soil Water Conserv. 44:168–173.
TOP
ABSTRACT
INTRODUCTION
