Estimating Root Zone Soil Moisture at Continental Scale Using Neural Networks

This study investigates the feasibility of artificial neural networks (ANNs) to retrieve root zone soil moisture (RZSM) at the depths of 20 cm (SM20) and 50 cm (SM50) at a continental scale, using surface information. To train the ANNs to capture interactions between land surface and various climatic patterns, data of 557 stations over the continental United States were collected. A sensitivity analysis revealed that the ANNs were able to identify input variables that directly affect the water and energy balance in root zone. The data important for RZSM retrieval in a large area included soil texture, surface soil moisture, and the cumulative values of air temperature, surface soil temperature, rainfall, and snowfall. The results showed that the ANNs had high skill in retrieving SM20 with a correlation coefficient above 0.7 in most cases, but were less effective at estimating SM50. The comparison of the ANNs showed that using soil texture data improved the model performance, especially for the estimation of SM50. It was demonstrated that the ANNs had high flexibility for applications in different climatic regions. The method was used to generate RZSM in North America using Soil Moisture and Ocean Salinity (SMOS) soil moisture data, and achieved a spatial soil moisture pattern comparable to that of Global Land Data Assimilation System Noah model with comparable performance to the SMOS surface soil moisture retrievals. The models can be efficient alternatives to assimilate remote sensing soil moisture data for shallow RZSM retrieval.     

STRATIFICATION OF VARIABILITY IN RUNOFF AND SEDIMENT YIELD BASED ON VEGETATION CHARACTERISTICS1

ABSTRACT: Runoff and sediment yield were collected from 100 plots during simulated rainfalls (100 mm/hr for 15 minutes) at antecedent soil moisture conditions. A clustering technique was used to stratify the variability of a single data set within a sagebrush‐grass community into four groups based on vegetation life form and amount of cover. The four cluster groups were grass, grass/shrub, shrub, and forb/grass and were found to be significantly different in plant height, surface roughness, soil bulk density, and soil organic matter. Stepwise multiple regression analyses were performed on the single data set and each cluster group. Results for individual groups resulted in more robust predictive equations for runoff (r²= 0.65–0.73) and sediment yield (r²= 0.37–0.91) than for equations developed from the single data set (r²= 0.56 for runoff and r²= 0.27 for sediment yield). The standard errors of the cluster group regression equations were also improved in three of the four group equations for both runoff and sediment yield compared to the single data set. Runoff was found to be significantly less (p >0.01) in the forb/grass group compared with other vegetation cluster groups, but this was influenced by four plots that produced little or no runoff. Sediment yield was not found to be significantly different among any cluster groups. Discriminant analysis was then used to identify important variables and develop a model to classify plots into one of the four cluster groups. The discriminant model could be incorporated into rangeland hydrology and erosion models. The percentage cover of grasses, shrubs, litter, and bare ground effectively stratified about 12 percent of the variation observed in runoff and 26 percent of the variability for sediment yield as determined by r².     

EVAPOTRANSPIRATION AND SOIL MOISTURE DYNAMICS ON A SEMIARID PONDEROSA PINE HILLSLOPE1

ABSTRACT: Detailed measurements of soil moisture and ET in semiarid forest environments have not been widely reported in the literature. In this study, soil moisture and water balance components were measured over a four‐year period on a semiarid ponderosa pine hillslope, with evapotranspiration (ET) determined as the residual of measured precipitation, runoff, and change in soil moisture storage. ET accounts for approximately 95 percent of the water budget and has a distinctly bimodal annual pattern, with peaks occurring after spring snowmelt and during the late summer monsoon season, periods that coincide with high soil moisture. Weekly growing season ET rates determined by the hillslope water balance are found to be invariably below calculated potential rates. Normalized ET rates are linearly correlated (r²= 0.62) with soil moisture; therefore, a simple linear relation is proposed. Growing season soil moisture dynamics were modeled based on this relation. Results are in fair agreement (r²= 0.63) with the observed soil moisture data over the four growing seasons; however, for two dry summers with little surface runoff, much better results (r² > 0.90) were obtained.     

Decoupling Streamflow Responses to Climate Variability and Land Use/Cover Changes in a Watershed in Northern China

Restored annual streamflow (Q_r) and measured daily streamflow of the Chaohe watershed located in northern China and associated long‐term climate and land use/cover data were used to explore the effects of land use/cover change and climate variability on the streamflow during 1961‐2009. There were no significant changes in annual precipitation (P) and potential evapotranspiration, whereas Q_r decreased significantly by 0.81 mm/yr (p < 0.001) over the study period with a change point in 1999. We used 1961‐1998 as the baseline period (BP) and 1999‐2009 the change period (CP). The mean Q_r during the CP decreased by 39.4 mm compared with that in the BP. From 1979 to 2009, the grassland area declined by 69.6%, and the forest and shrublands increased by 105.4 and 73.1%, respectively. The land use/cover change and climate variability contributed for 58.4 and 41.6% reduction in mean annual Q_r, respectively. Compared with the BP, median and high flows in the CP decreased by 38.8 and up to 75.5%, respectively. The study concludes that large‐scale ecological restoration and watershed management in northern China has greatly decreased water yield and reduced high flows due to the improved land cover by afforestation leading to higher water loss through evapotranspiration. At a large watershed scale, land use/cover change could play as much of an important role as climate variability on water resources.     

Neural Network Input Selection for Hydrological Forecasting Affected by Snowmelt1

Abstract: Snowmelt largely affects runoff in watersheds in Nordic countries. Neural networks (NN) are particularly attractive for streamflow forecasting whereas they rely at least on daily streamflow and precipitation observations. The selection of pertinent model inputs is a major concern in NNs implementation. This study investigates performance of auxiliary NN inputs that allow short‐term streamflow forecasting without resorting to a deterministic snowmelt routine. A case study is presented for the Rivière des Anglais watershed (700 km²) located in Southern Québec, Canada. Streamflow (Q), precipitations (rain R and snow S, or total P), temperature (T) and snow lying (A) observations, combined with climatic and snowmelt proxy data, including snowmelt flow (Q_SM) obtained from a deterministic model, were tested. NN implemented with antecedent Q and R produced the largest gains in performance. Introducing increments of A and T to the NNs further improved the performance. Long‐term averages, seasonal data, and Q_SM failed to improve the networks.     

Passive aeration composting of chicken litter: Effects of aeration pipe orientation and perforation size on losses of compost elements

A passive aeration composting study was undertaken to investigate the effects of aeration pipe orientation (PO) and perforation size (PS) on some physico-chemical properties of chicken litter (chicken manure + sawdust) during composting. The experimental set up was a two-factor completely randomised block design with two pipe orientations: horizontal (Ho) and vertical (Ve), and three perforation sizes: 15, 25 and 35 mm diameter. The properties monitored during composting were pile temperature, moisture content (MC), pH, electrical conductivity (EC), total carbon (C_T), total nitrogen (N_T) and total phosphorus (P_T). Moisture level in the piles was periodically replenished to 60% for efficient microbial activities. The results of the study showed that optimum composting conditions (thermophilic temperatures and sanitation requirements) were attained in all the piles. During composting, both PO and PS significantly affected pile temperature, moisture level, pH, C_T loss and P_T gain. EC was only affected by PO while N_T was affected by PS. Neither PO nor PS had a significant effect on the C:N ratio. A vertical pipe was effective for uniform air distribution, hence, uniform composting rate within the composting pile. The final values showed that PO of Ve and PS of 35 mm diameter resulted in the least loss in N_T. The PO of Ho was as effective as Ve in the conservation of C_T and P_T. Similarly, the three PSs were equally effective in the conservation of C_T and P_T. In conclusion, the combined effects of PO and PS showed that treatments Ve35 and Ve15 were the most effective in minimizing N_T loss.     

Evaluating the BANCS Streambank Erosion Framework on the Northern Gulf of Mexico Coastal Plain

The Bank Assessment of Nonpoint source Consequences of Sediment (BANCS) framework allows river scientists to predict annual sediment yield from eroding streambanks within a hydrophysiographic region. BANCS involves field data collection and the calibration of an empirical model incorporating a bank erodibility hazard index (BEHI) and near‐bank shear stress (NBS) estimate. Here we evaluate the applicability of BANCS to the northern Gulf of Mexico coastal plain, a region that has not been previously studied in this context. Erosion rates averaged over two years expressed the highest variability of any existing BANCS study. As a result, four standard BANCS models did not yield statistically significant correlations to measured erosion rates. Modifications to two widely used NBS estimates improved their correlations (r² = 0.31 and r² = 0.33), but further grouping of the data by BEHI weakened these correlations. The high variability in measured erosion rates is partly due to the regional hydrologic and climatic characteristics of the Gulf coastal plains, which include large, infrequent precipitation events. Other sources of variability include variations in bank vegetation and the complex hydro‐ and morphodynamics of meandering, sand bed channels. We discuss directions for future research in developing a streambank erosion model for this and similar regions.     

Soil Properties Predict Plant Community Development of Mitigation Wetlands Created in the Virginia Piedmont,USA

The study investigated vegetative and soil properties in four created mitigation wetlands, ranging in age from three to ten years, all created in the Virginia Piedmont. Vegetation attributes included percent cover, richness (S), diversity (H′), floristic quality assessment index (FQAI), prevalence index (PI), and productivity [i.e., peak above-ground biomass (AGB) and below-ground biomass]. Soil attributes included soil organic matter (SOM), gravimetric soil moisture (GSM), pH, and bulk density (D_b) for the top 10 cm. Species dominance (e.g., Juncus effusus, Scirpus cyperinus, Arthraxon hispidus) led to a lack of differences in vegetative attributes between sites. However, site-based differences were found for GSM, pH, and SOM (P < 0.001). Soil attributes were analyzed using Euclidean cluster analysis, resulting in four soil condition (SC) categories where plots were grouped based on common attribute levels (i.e., SC1 > SC2 > SC3 > SC4, trended more to less developed). When vegetation attributes were compared between SC groups, greater SOM, lower D_b, more circumneutral pH, and higher GSM, all indicative of maturation, were associated with higher H′ (P < 0.05), FQAI (P < 0.05), and total and volunteer percent cover (P < 0.05), and lower AGB (P < 0.001), PI (P < 0.05), and seeded percent cover (P < 0.05). The outcome of the study shows that site age does not necessarily equate with site development with soil and vegetation developmental rates varying both within and among sites. The inclusion of soil attributes in post-construction monitoring should be required to enhance our understanding and prediction of developmental trajectory of created mitigation wetlands.     

THE EFFECTS OF A MOVING RAINSTORM ON DIRECT RUNOFF PROPERTIES1

ABSTRACT: The effects of a moving rainstorm on flood runoff characteristics were investigated. A flood hydrograph simulation model called “FH-Model” and a natural watershed were used. A hypothetical rainstorm of 50 years recurrence interval, 75 mm depth, and 4 hours duration was used to show the effects of velocity and direction of the moving rainstorm on the runoff characteristics. Compared with an equivalent stationary rainstorm (ESRS), the peak flow caused by a rainstorm moving in a downstream direction with a speed equal to channel velocity, V, was 27.5 percent higher and the peak flow caused by the same rainstorm moving in an upstream direction was 21.7 percent smaller. These percentages reduced to 10.5 percent and 8.6 percent for storms moving downstream and upstream, respectively, at three times the channel velocity, 3V. There were negligible differences in the time of peak, T_p between runoff caused by storms moving downstream and runoff produced by ESRS. However, T_p for a storm moving upstream at V velocity was 82 percent higher than that produced by ESRS, but was reduced to 27 percent higher when the storm velocity was 3V.     

Evaluation of the European Space Agency Climate Change Initiative Soil Moisture Product over China Using Variance Reduction Factor

In this study, we evaluated the European Space Agency Climate Change Initiative soil moisture product v02.1 (ESA CCI SM v02.1) using in situ observations collected at 547 stations in China from 1991 to 2013. A conventional validation was first conducted, and the triple collocation errors of ESA CCI SM and the European Centre for Medium Range Weather Forecasting reanalysis data were approximately 0.053 and 0.050 m³/m³, respectively. To obtain more reliable validation results, the average soil moisture of the in situ observations per ESA CCI SM pixel was also used as the validation sites. Variance reduction factor (VRF) was adopted to quantify the accuracy of the soil moisture validation sites, and the average VRF was estimated at 4.88%. The validation results were enhanced by excluding validation sites with VRF errors greater than 5% from the statistical analysis. Although the ESA CCI SM underestimated the in situ observations with a Bias of 0.05 m³/m³, a moderately high correlation coefficient of 0.44 and a relatively small unbiased root‐mean‐square difference of 0.05 m³/m³ were observed. This study provides information on the utilization of ESA CCI SM for ecological protection, climate change, and hydrological forecasting. It also suggests the adoption of VRF for future error corrections of satellite‐based products.     

STREAM NITRATE-N LOADS IN RELATION TO VARIATIONS IN ANNUAL AND SEASONAL RUNOFF REGIMES1

ABSTRACT: The calculation of stream nutrient loads from a sampling period of one year or, at most, a few years may provide an inaccurate estimate of average seasonal or annual loads due to considerable year-to-year variations in hydrological regime. The number of years of record required to give a reliable estimate of long-term average NO₃-N loads was analyzed for E. Duffin Creek and the Nottawasaga River in Ontario, Canada. Nitrate load rating relationships were used in combination with a continuous stream discharge record for 22 years (E. Duffin Creek) and 34 years (Nottawasaga River) to simulate long-term seasonal and annual variation in NO₃.N loads. The errors involved in calculating average loads were examined by comparing the loads derived from sampling periods of one or more consecutive years duration with the estimated long-term average load for the two rivers. Annual NO₃-N loads for a single year deviated from the long-term average load by ± 20 to 53 percent in 8 out of 22 years in E. Duffin Creek and in 13 of 34 years in the Nottawasaga River. Six consecutive years of record would be required for both rivers to ensure that an error of > ± 20 percent would occur in only 5 percent of these observation periods. February-April NO₃-N loads for a single year could deviate by up to +90 percent or -61 percent from the long-term average spring period load for the two rivers. A sampling period of at least 6–7 years would be needed to estimate average NO₃-N loads for the spring runoff season with an error <± 20 percent.     

CLIMATE CHHANGE SENSITIVITY ASSESSMENT ON UPPER MISSISSIPPI RIVER BASIN STREAMFLOWS USING SWAT1

ABSTRACT: The Soil and Water Assessment Tool (SWAT) model was used to assess the effects of potential future climate change on the hydrology of the Upper Mississippi River Basin (UMRB). Calibration and validation of SWAT were performed using monthly stream flows for 1968–1987 and 1988–1997, respectively. The R² and Nash‐Sutcliffe simulation efficiency values computed for the monthly comparisons were 0.74 and 0.69 for the calibration period and 0.82 and 0.81 for the validation period. The effects of nine 30‐year (1968 to 1997) sensitivity runs and six climate change scenarios were then analyzed, relative to a scenario baseline. A doubling of atmospheric CO₂ to 660 ppmv (while holding other climate variables constant) resulted in a 36 percent increase in average annual streamflow while average annual flow changes of ?49, ?26, 28, and 58 percent were predicted for precipitation change scenarios of ?20, ?10, 10, and 20 percent, respectively. Mean annual streamflow changes of 51,10, 2, ?6, 38, and 27 percent were predicted by SWAT in response to climate change projections generated from the CISRO‐RegCM2, CCC, CCSR, CISRO‐Mk2, GFDL, and HadCMS general circulation model scenarios. High seasonal variability was also predicted within individual climate change scenarios and large variability was indicated between scenarios within specific months. Overall, the climate change scenarios reveal a large degree of uncertainty in current climate change forecasts for the region. The results also indicate that the simulated UMRB hydrology is very sensitive to current forecasted future climate changes.     

The Potential Use of Synthetic Aperture Radar for Estimating Methane Ebullition From Arctic Lakes1

Abstract: Arctic lakes are significant emitters of methane (CH₄), a potent greenhouse gas, to the atmosphere; yet no rigorous quantification of the magnitude and variability of pan‐Arctic lake emissions exists. In this study, we demonstrate the potential for a new method using synthetic aperture radar (SAR) imagery to detect methane bubbles in lake ice to scale up whole‐lake measurements of CH₄ ebullition (bubbling) to regional scales. We estimated ebullition from lakes, which is often the dominant mode of lake emissions, by mapping the distribution of bubble clusters frozen in early winter ice across surfaces of seven tundra lakes and one boreal forest lake in Alaska. Applying previously measured ebullition rates associated with four distinct classes of bubble clusters found in lake ice, we estimated whole‐lake emissions from individual lakes. The percent surface area of lake ice covered with bubbles (R² = 0.68) and CH₄ ebullition rates from lakes (R² = 0.59) and were correlated with radar return values from RADARSAT‐1 Standard Beam mode 3 for the tundra lakes, suggesting that with appropriate scaling and consideration for variability in lake‐ice conditions, this technique has the potential to be used for estimating broader‐scale regional and pan‐Arctic lake methane emissions.     

IMPROVED RAINFALL/RUNOFF ESTIMATES USING REMOTELY SENSED SOIL MOISTURE1

ABSTRACT: Remotely sensed soil moisture data measured during the Southern Great Plains 1997 (SGP97) experiment in Oklahoma were used to characterize antecedent soil moisture conditions for the Soil Conservation Service (SCS) curve number method. The precipitation‐adjusted curve number and the soil moisture were strongly related (r²= 0.70). Remotely sensed soil moisture fields were used to adjust the curve numbers and the runoff estimates for five watersheds, in the Little Washita watershed; the results ranged from 2.8 km² to 601.6 km². The soil moisture data were applied at two spatial scales, a finer one (800 m) measuring spatial resolution and a coarser one (28 km). The root mean square error (RMSE) and the mean absolute error (MAE) of the runoff estimated by the standard SCS method was reduced by nearly 50 percent when the 800 m soil moisture data were used to adjust the curve number. The coarser scale soil moisture data also significantly reduced the error in the runoff predictions with 41 percent and 28 percent reductions in MAE and RMSE, respectively. The results suggest that remote sensing of soil moisture, when combined with the SCS method, can improve rainfall runoff predictions at a range of spatial scales.     

PHOSPHORUS LOSSES FROM CROPLAND AS AFFECTED BY TILLAGE SYSTEM AND FERTILIZER APPLICATION METHOD1

ABSTRACT: A rainfall simulator was used to study the effectiveness of no-till and fertilizer application method on reducing phosphorus (P) losses from agricultural lands. Simulated rainfall was applied to 12 experimental field plots, each 0.01 ha in size. The plots were divided into no-till and conventional tillage systems. Two fertilizer application methods, subsurface injection and surface application, were investigated for the two tillage systems. Phosphorus fertilizer was applied at a rate of 46 kg/ha, 24 to 48 hours before the start of rain simulation. Water samples were collected from the base of each plot and analyzed for sediment and P content. No-till was found to be very effective in reducing runoff and sediment losses. No-till reduced sediment loss and total runoff volume by 92 and 67 percent, respectively. Subsurface injection of fertilizer, as compared to surface application, reduced PO₄ losses by 39 percent for no-till and by 35 percent for conventional tillage. The effect of tillage system on PO₄ losses was not significant. Reductions in total-P (P_T) losses due to no-till compared to the conventional tillage system were 89 and 91 percent for surface application and subsurface injection methods, respectively. Averaged across all fertilizer treatments, an equivalent of 0.9 and 8.9 percent of the P applied to the plots were lost from the no-till and conventional tillage plots, respectively.     

OPTICAL HETEROGENEITY IN GREEN BAY1

Differences in light penetration and light attenuating components and processes are documented along 112 km of the major (NE/SW) axis of Green Bay (Lake Michigan) during a three-day cruise (May 25–27, 1982). Measurements included diffuse attenuation of downwelling irradiance (k_d), Secchi disk transparency (SD), phytoplankton pigments (chlorophyll a and phaeophytin), turbidity (T_n), and dissolved color (absorbance). The relative importance of absorption and scattering to attenuation was calculated from paired measurements of k_d and SD. Absorption and scattering coefficients were calculated; value estimates were supported by a strong linear relationship between the scattering coefficient (b) and T_n (b = 0.99 *T_n; r²= 097, n = 12). Light attenuation characteristics, including the extent of light penetration and the magnitudes and relative importance of absorption, scattering and individual attenuating components, were found to be heterogenous in space. This heterogeneity is due to the characteristics and positions of entry of fluvial discharges to the bay as they influence levels of dissolved color (Gelbstoff), phytoplankton standing crop, and inorganic particulates. Identification of key processes regulating light penetration and their potential for response to pollution control measures can aid in the development of a water quality management plan for Green Bay.     

Soil and plant characteristics of landfill sites near Merseyside,England

An ecological survey of the plant and soil characteristics was carried out on three landfill sites near Merseyside, England. It was discovered that bare ground at two of the landfill areas had high levels of methane contained in the soil air (Sefton Meadows landfill: 6–8% at 35 cm and 16–35% at 65 cm below soil surface; Coalgate Lane landfill: 1–24% at 35 cm and 39–45% at 40 cm below soils surface), causing the appearance of dark grey reduced regions in the soil, a phenomenon similar to flooded soil. The wellvegetated areas at the two sites had lower levels of methane (under 7%).In areas relatively free of methane, the concentrations of mineralized N and NO₃ ^– had significant correlations with the dry weights of vegetation (r = 0.71 withp<0.01;r=0.61 withp<0.02 accordingly), indicating the necessity of applying available nitrogen fertilizer.     

Calibrating a Basin‐Scale Groundwater Model to Remotely Sensed Estimates of Groundwater Evapotranspiration

Remotely sensed vegetation indices correspond to canopy vigor and cover and have been successfully used to estimate groundwater evapotranspiration (ET_g) over large spatial and temporal scales. However, these data do not provide information on depth to groundwater (dtgw) necessary for groundwater models (GWM) to calculate ET_g. An iterative approach is provided that calibrates GWM to ET_g derived from Landsat estimates of the Enhanced Vegetation Index (EVI). The approach is applied to different vegetation groups in Mason Valley, Nevada over an 11‐year time span. An uncertainty analysis is done to estimate the resulting mean and 90% confidence intervals in ET_g to dtgw relationships to quantify errors associated with plant physiologic complexity, species variability, and parameter smoothing to the 100 m GWM‐grid, temporal variability in soil moisture and nonuniqueness in the solution. Additionally, a first‐order second moment analysis shows ET_g to dtgw relationships are almost exclusively sensitive to estimated land surface, or maximum, ET_g despite relatively large uncertainty in extinction depths and hydraulic conductivity. The EVI method of estimating ET_g appears to bias ET_g during years with exceptionally wet spring/summer conditions. Excluding these years improves model performance significantly but highlights the need to develop a methodology that accounts not only on quantity but timing of annual precipitation on phreatophyte greenness.     

Investigation of the Curve Number Method For Surface Runoff Estimation In Tropical Regions

This study tests the applicability of the curve number (CN) method within the Soil and Water Assessment Tool (SWAT) to estimate surface runoff at the watershed scale in tropical regions. To do this, surface runoff simulated using the CN method was compared with observed runoff in numerous rainfall‐runoff events in three small tropical watersheds located in the Upper Blue Nile basin, Ethiopia. The CN method generally performed well in simulating surface runoff in the studied watersheds (Nash‐Sutcliff efficiency [NSE] > 0.7; percent bias [PBIAS] < 32%). Moreover, there was no difference in the performance of the CN method in simulating surface runoff under low and high antecedent rainfall (PBIAS for both antecedent conditions: ~30%; modified NSE: ~0.4). It was also found that the method accurately estimated surface runoff at high rainfall intensity (e.g., PBIAS < 15%); however, at low rainfall intensity, the CN method repeatedly underestimated surface runoff (e.g., PBIAS > 60%). This was possibly due to low infiltrability and valley bottom saturated areas typical of many tropical soils, indicating that there is scope for further improvements in the parameterization/representation of tropical soils in the CN method for runoff estimation, to capture low rainfall‐intensity events. In this study the retention parameter was linked to the soil moisture content, which seems to be an appropriate approach to account for antecedent wetness conditions in the tropics.     

GOES IR AND DUMMY VARIABLE TECHNIQUES FOR LOW TEMPERATURE ASSESSMENT IN FLORIDA1

ABSTRACT: Geostationary Operational Environmental Satellite (GOES) thermal infrared (IR) data were used to study regional cold nocturnal temperature fluctuations that are important in assessing citrus freeze damage. Dummy variables techniques were used to analyze the temperature difference between the east and west sides of Central Florida, which was obtained from both ground measurement minimum temperature (T_GMM) and the GOES mean temperature (T_GOES) on low-temperature nights. The low temperature pattern, which was closely related to the citrus freeze damage pattern in Central Florida, was identified by GOES thermal IR data but not by conventional ground measurements. The low surface temperature in the west side and east side appears to be attributable to differences in soil types, water tables, and drainage classes in both regions.