Jene Michaud

Comparison of simple versus complex distributed runoff models on a midsized semiarid watershed

The increasing availability of distributed rainfall data and computational resources is providing the opportunity to use distributed models for rainfall-runoff forecasting or other applications. This paper compares the accuracy of simulations from a complex distributed model (KINEROS), a simple distributed model (based on the Soil Conservation Service (SCS) method), and a simple lumped model (SCS method). The 150 km2, semiarid Walnut Gulch experimental watershed was the test site; models were validated using 24 severe thunderstorms and rain gauge densities similar to those found at flash flood warning sites (one gauge per 20 km2). Under these circumstances, none of the models were able to accurately simulate peak flows or runoff volumes from individual events. Models showed somewhat more skill in predicting time to peak and the ratio of peak flow to volume. When calibration was performed, the accuracy of the complex distributed model was similar to that of the simple distributed model. Without calibration, the complex distributed model was more accurate than the simple distributed model. The spatially lumped model performed very poorly. The complex distributed model was validated under real-time forecasting conditions; forecasts based on observed rainfall had lead times of 30—75 min.

Effect of rainfall-sampling errors on simulations of desert flash floods

The effect of rainfall-sampling errors on distributed hydrologic simulations was evaluated in a study conducted with localized thunderstorms and a midsized (150 km2) semiarid watershed. Rainfall fields based on observations from a very dense rain gage network were compared to rainfall fields based on observations from a subset of the original gages. The rain gage density of the “sparse” network (1 gage per 20 km2) was selected to represent the typical gage density of a local evaluation in real time (ALERT) type flash flood warning system. Inadequate rain gage densities in the case of the sparse network produced errors in simulated peaks that, on the average, represented 58% of the observed peak flow. Approximately half of the difference between observed and simulated peaks was due to rainfall-sampling errors. Simulations were also conducted with rainfall that is similar to the next generation weather radar (NEXRAD) digital precipitation estimates in that it represents areal averages within 4 km × 4 km pixels. Spatial averaging of rainfall over 4 km × 4 km pixels led to consistent reductions in simulated peaks that, on the average, represented 50% of the observed peak flow. Hence it appears that the current spatial resolution of ALERT-type precipitation measurements and 4 km × 4 km radar precipitation estimates may not be sufficient to produce reliable rainfall-runoff simulations/forecasts in midsized watersheds of the southwestern United States subject to localized thunderstorms and large infiltration losses.

Ecosystem and Restoration Consequences of Invasive Woody Species Removal in Hawaiian Lowland Wet Forest

A removal experiment was used to examine the restoration potential of a lowland wet forest in Hawaii, a remnant forest type that has been heavily invaded by non-native species and in which there is very little native species regeneration. All non-native woody and herbaceous biomass (approximately 45% of basal area) was removed in four 100-m2 removal plots; plots were followed for a three-year period. Removal plots had a lower leaf area index, higher air temperatures, higher afternoon soil temperatures, and lower relative humidity than control plots. Removal plots had 40% less litterfall mass and similarly reduced nutrient inputs. Leaf litter decomposition rates were much slower in the removal plots, due more to site quality than litter quality. However, soil N and P were not different between treatments. Native species had a distinct suite of leaf traits (greater integrated water use efficiency, lower mass-based leaf nutrient concentrations, and lower specific leaf area). Despite major environmental changes in the removal plots, native species’ diameter growth and litterfall productivity were not significantly greater after removal, testifying to the slow response capabilities of native Hawaiian trees. Our results are consistent with the expectation that native species are conservative in regards to resource use and may not strongly respond to canopy removal, at least at the adult stage. Management strategies will have to incorporate the slow growth rate of Hawaiian species and the fact that weeding may be required to suppress expansion and nutrient inputs of introduced species.

Testing of vegetation parameter aggregation rules applicable to the Biosphere-Atmosphere Transfer Scheme (BATS) and the FIFE site

Abstract A realistic model of surface-atmosphere exchanges was created by coupling the Biosphere-Atmosphere Transfer Scheme (BATS) with an advanced, two-dimensional model of the atmospheric boundary layer. This was initiated and tested using data obtained from the First International Satellite Land Surface Climatology Project Field Experiment (FIFE). This model was used to investigate the acceptability of simple rules for defining the aggregate value of the parameters required to specify surface interactions, as applied to heterogeneous mixes of vegetation types allowed in BATS but appropriate to the FIFE site, namely short and long grass, mixed crops, and irrigated crops. Under the range of meteorological and surface conditions relevant to FIFE and as used in this study, these rules are shown to estimate aggregate parameters which give surface fluxes similar to those calculated with an explicit representation of separate vegetation patches, except in the particular case of artificially wetted (irrigated) patches set in an otherwise dry landscape.

Regional variations in small-basin floods in the United States

This paper presents a reconnaissance study of regional flood patterns in the United States, focusing on peak discharges at several recurrence intervals and characteristics of flood-causing rainfall. Because of an interest in flash floods, attention was restricted to watersheds between 10 and 200 km 2 in area. Data were obtained from 130 stream gaging stations with a consistent 30-year period of record and from reports documenting 90 exceptional floods occurring mostly on ungaged watersheds. Peak discharges vary considerably within local regions. Roughly 60% of the local variability can be explained by watershed characteristics, but watershed area is not a reliable predictor of peak discharge within the narrow range of watershed sizes examined. On a continental scale the spatial patterns of the median and 25-year floods are similar. In both cases a concentration of large floods is found in the southeastern Great Plains and parts of the southeast. In the west, north, and northeast, floods tend to be small, but large floods still occur in scattered locations. The pattern and seasonality of the exceptional floods, which are presumed to have relatively long recurrence intervals, are different from the pattern of median and 25-year floods. The largest of the exceptional floods are concentrated in the central and southern Great Plains during May and June. They occur farther west (and several months later) than the largest median floods. Exceptional floods occurring in the semiarid west were caused by as little as 5-10 cm of rain in 30 - 60 min, whereas in humid areas most of the exceptional floods resulted from 13-32 cm of rain in 1-12 hours.

Executive summary of the Tucson Aggregation Workshop

Abstract ‘Aggregation’ refers to spatial averaging of some heterogeneous surface variable to obtain an effective value representative of an area. The effect of surface heterogeneity on interactions between land and atmosphere is relevant to near-surface hydrology, ecology, and climate, and is the common theme of the papers in this issue. Even though the full effect of heterogeneity must be neglected owing to limited spatial resolution of large-scale models, it is important to understand when and how the presence of heterogeneity requires recognition in any aggregate representation. In March 1994, a workshop, which has come to be known as the ‘Tucson Aggregation Workshop’, was convened to assess the state of the art in aggregation research, and the papers in this issue are the product of that workshop. The principal findings of the workshop can be summarized as follows: 1. 1. substantial progress has been made in producing aggregated representations of flat terrain. Simple aggregation rules applied to surface properties have given rise, in some studies, to simulated surface energy fluxes that are within 10% of fluxes produced from models with full representation of heterogeneity. 2. 2. Aggregation rules are relatively straightforward in the case of patch-scale heterogeneity (variability on the order of hundreds to thousands of meters) of vegetative characteristics which control surface exchanges, although aggregation of soil hydraulic properties and possibly of soil moisture remains problematic. In addition, some of the effects of meso-scale heterogeneity (variability on the order or 10–100 km) in surface cover will need to be addressed through more complicated types of parameterization. 3. 3. There is convincing evidence that the regional energy balance (over, say, 10 5 km 2 ) is insensitive to gentle topography, provided that surface vegetation and water availability are uniform, but in mountainous terrain the influence of topography on near-surface meteorology must be considered. 4. 4. It appears that the value of simple combinations of remotely sensed radiances representing areal-average measurements are influenced only slightly by unresolved variability, although the averaging of some derived variables based on these radiances offers a greater challenge, especially with sparse canopies.

Persistence of Native Trees in an Invaded Hawaiian Lowland Wet Forest: Experimental Evaluation of Light and Water Constraints

Abstract: Hawaiian lowland wet forests are heavily invaded and their restoration is most likely to be successful if native species selected for restoration have efficient resource-use traits. We evaluated growth, survival, and ecophysiological responses of four native and four invasive species in a greenhouse experiment that simulated reduced light and water conditions commonly found in invaded field conditions. Our results show that light is a more important limiting resource than water for all species. Specifically, values for photosynthesis, light compensation point, light saturation point, stomatal conductance, leaf mass per area, relative growth rate, and photosynthetic nitrogen use efficiency were all greater under high-light conditions than they were under low-light conditions. In contrast, water limitation negatively affected only stomatal conductance and &dgr;13C. Our results also show that responses to light were species-specific rather than related to whether species were native or nonnative. We also tested restoration potential of top-performing native species under field conditions in a Hawaiian lowland wet forest by comparing relative growth and mortality rates in both invaded (low-light) plots and in plots from which invasive species had been removed (high-light conditions). Of the native species, Myrsine lessertiana and Psychotria hawaiiensis had highest survival and growth rates in low-light plots after 4 yr, and Metrosideros polymorpha showed 100% mortality under the same conditions. Under low light, M. lessertiana and P. hawaiiensis survived and grew at rates similar to those of invasive species in both field and greenhouse and thus represent suitable candidates for restoration in invaded Hawaiian lowland wet forests.

Drought in an invaded Hawaiian lowland wet forest.

Abstract: In this study we examined historic drought frequency and hydrologic effects of removing invasive plants from one of the few remaining Hawaiian wet lowland forests, near Hilo, Hawai‘i. We developed a conceptual and statistical model of Hilo droughts using historic rainfall and pan evaporation data and discovered that episodes of low soil moisture were most likely from January to March but also occurred in June or July. Field measurements were taken in four pairs of plots. Nonnative woody and herbaceous species were removed from four plots; control plots were undisturbed. Soil water potential measurements documented partial soil drying in control plots, but not removal plots, during droughts with recurrence intervals of 2–3 yr. Drier soils exhibited strong small-scale heterogeneity in soil water potential that presumably reflects macroporosity in the young ‘a‘ā lava flow substrate. Transpiration from and rainfall interception by the dense canopy of nonnative species were most likely responsible for drier conditions in control plots. Removal plots experienced changes to shading, midday vapor pressure deficit, albedo, and aerodynamic resistance, but it appears that hydrologic impact of these variables was minor. We suggest that efforts to restore Hawaiian tropical rain forests should consider drought resilience as one objective, among many, of a restoration program. Germinating seeds, shallowrooted saplings and deeper-rooted mature trees may respond differently to hydrologic effects of removing invasive plants.

Dam Break Inundation Study for the State of Hawaii

This paper will describe a dam break inundation study that was performed for the State of Hawaiis Department of Land and Natural Resources (DLNR). The purpose of the study was to produce updated flood inundation maps for emergency preparedness and evacuation planning throughout the state, as well as to provide the basis for a hazard risk assessment for the downstream area. A total of 135 dams were studied and mapped. Dam breaks were modeled as single reservoir failures and also as sequential failures. The flood inundation maps were produced using a combination of 1 and 2-dimensional numerical hydrodynamic models. The paper will discuss the benefits of using a 2D hydraulic model compared to the classical approach of using a 1D hydraulic model for dam break flood inundation mapping. Pacific Disaster Center performed the study with the assistance of the University of Hawaii at Hilo and Danish Hydraulic Institute.

Methods for Estimating the Impact of Hypothetical Dam Break Floods

This paper presents a method for estimating the impact of floods resulting from dam failure. These methods were developed for implementation at the Pacific Disaster Center in Hawaii. A hydraulic model embedded in a Geographic Information System (GIS) is used to estimate the downstream attenuation of the flood. The model output is then integrated with a Digital Elevation Model in order to estimate the lateral extent and depth of flooding. Information useful for disaster planning is obtained by overlaying the flood map onto GIS infrastructure layers (roads, emergency services, and chemical plants).