Search Results

You are looking at 1 - 10 of 18 items for

  • Author or Editor: P. C. D. Milly x
  • Refine by Access: All Content x
Clear All Modify Search
P. C. D. Milly

Abstract

The parameterization of continental evaporation in many atmospheric general circulation models (GCMS) used for simulation of climate is demonstrably inconsistent with the empirical work upon which the parameterization is based. In the turbulent transfer relation for potential evaporation, the climate models employ the modeled actual temperature to evaluate the saturated surface humidity, whereas the consistent temperature is the one reflecting cooling by the hypothetical potential evaporation. A simple theoretical analysis and some direct computations, all ignoring atmospheric feedbacks, indicate that whenever the soil moisture is limited, GCM-based climate models produce rates of potential evaporation that exceed, by a factor of two or more, the rates that would be yielded by use of the consistent temperature. Further approximate analyses and supporting numerical simulations indicate that the expected value of dry-season soil moisture has a short memory relative to the annual cycle and that dry-season evaporation is therefore nearly equal to dry-season precipitation. When potential evaporation is overestimated, it follows that the soil moisture is artificially reduced by a similar factor, and actual evaporation may or may not be overestimated, depending on other details of the hydrologic parameterization. These arguments, advanced on theoretical grounds, explain the substantial, systematic differences between GCM-generated and observation-based estimates of potential evaporation rates and call into question the direct use of currently available GCM-generated values of potential evaporation in the assessment of the effects of climatic change on continental hydrology and water resources. They also provide a partial explanation of the excessively low values of summer soil moisture in GCMs and raise questions concerning the results of studies of soil-moisture changes induced by an increase of greenhouse gases. Nevertheless, an approximate analytical result suggests that the basic dependence of changes in soil moisture on changes in the atmospheric state was qualitatively preserved in those studies.

Full access
Randal D. Koster and P. C. D. Milly

Abstract

The Project for Intercomparison of Land-surface Parameterization Schemes (PILPS) has shown that different land surface models (LSMs) driven by the same meteorological forcing can produce markedly different surface energy and water budgets, even when certain critical aspects of the LSMs (vegetation cover, albedo, turbulent drag coefficient, and snowcover) are carefully controlled. To help explain these differences, the authors devised a monthly water balance model that successfully reproduces the annual and seasonal water balances of the different PILPS schemes. Analysis of this model leads to the identification of two quantities that characterize an LSM’s formulation of soil water balance dynamics: 1) the efficiency of the soil’s evaporation sink integrated over the active soil moisture range, and 2) the fraction of this range over which runoff is generated. Regardless of the LSM’s complexity, the combination of these two derived parameters with rates of interception loss, potential evaporation, and precipitation provides a reasonable estimate for the LSM’s simulated annual water balance. The two derived parameters shed light on how evaporation and runoff formulations interact in an LSM, and the analysis as a whole underscores the need for compatibility in these formulations.

Full access
C. Adam Schlosser and P. C. D. Milly

Abstract

Soil moisture predictability and the associated predictability of continental climate are explored as an initial-value problem, using a coupled land–atmosphere model with prescribed ocean surface temperatures. Ensemble simulations are designed to assess the extent to which initial soil moisture fields explain variance of future predictands (soil moisture, near-surface air temperature, and precipitation). For soil moisture, the decrease of explained variance with lead time can be characterized as a first-order decay process, and a predictability timescale is introduced as the lead time at which this decay reaches e−1. The predictability timescale ranges from about 2 weeks or less (in midlatitudes during summer, and in the Tropics and subtropics) to 2–6 months (in mid- to high latitudes for simulations that start in late fall and early winter). The predictability timescale of the modeled soil moisture is directly related to the soil moisture's autocorrelation timescale. The degree of translation of soil moisture predictability to predictability of any atmospheric variable can be characterized by the ratio of the fraction of explained variance of the atmospheric variable to the fraction of explained soil moisture variance. By this measure, regions with the highest associated predictability of 30-day-mean near-surface air temperature (ratio greater than 0.5) are, generally speaking, coincident with regions and seasons of the smallest soil moisture predictability timescales. High associated temperature predictability is found where strong variability of soil moisture stress on evapotranspiration and abundant net radiation at the continental surface coincide. The associated predictability of 30-day-mean precipitation, in contrast, is very low.

Full access
P. C. D. Milly and K. A. Dunne

Abstract

The sensitivity of the global water cycle to the water-holding capacity of the plant-root zone of continental soils is estimated by simulations using a mathematical model of the general circulation of the atmosphere, with prescribed ocean surface temperatures and prescribed cloud. With an increase of the globally constant storage capacity, evaporation from the continents rises and runoff falls, because a high storage capacity enhances the ability of the soil to store water from periods of excess for later evaporation during periods of shortage. In addition to this direct effect, atmospheric feedbacks associated with the resulting higher precipitation and lower potential evaporation drive further changes in evaporation and runoff. Most of the changes in evaporation and runoff occur in the tropics and in the northern middle-latitude rain belts. Global evaporation from land increases by about 7 cm for each doubling of storage capacity in the range from less than 1 cm to almost 60 cm. Sensitivity is negligible for capacity above 60 cm.

In the tropics and in the extratropics, the increased continental evaporation is split, in approximately equal parts, between increased continental precipitation and decreased convergence of atmospheric water vapor from ocean to land. In the tropics, this partitioning is strongly affected by induced circulation changes, which are themselves forced by changes in latent beating. The increased availability of water at the continental surfaces leads to an intensification of the Hadley circulation and a weakening of the monsoonal circulations. In the northern middle and high latitudes, the increased continental evaporation moistens the atmosphere. This change in humidity of the atmosphere is greater above the continents than above the oceans, and the resulting reduction in the sea-land humidity gradient causes a decreased onshore transport of water vapor by transient eddies.

Results established here may have implications for certain problems in global hydrology and climate dynamics, including the effects of water resource development on global precipitation, climatic control of plant rooting characteristics, climatic effects of tropical deforestation, and climate-model errors induced by errors in land-surface hydrologic parameterizations.

Full access
P. C. D. Milly and A. B. Shmakin

Abstract

Land water and energy balances vary around the globe because of variations in amount and temporal distribution of water and energy supplies and because of variations in land characteristics. The former control (water and energy supplies) explains much more variance in water and energy balances than the latter (land characteristics). A largely untested hypothesis underlying most global models of land water and energy balance is the assumption that parameter values based on estimated geographic distributions of soil and vegetation characteristics improve the performance of the models relative to the use of globally constant land parameters. This hypothesis is tested here through an evaluation of the improvement in performance of one land model associated with the introduction of geographic information on land characteristics. The capability of the model to reproduce annual runoff ratios of large river basins, with and without information on the global distribution of albedo, rooting depth, and stomatal resistance, is assessed. To allow a fair comparison, the model is calibrated in both cases by adjusting globally constant scale factors for snow-free albedo, non-water-stressed bulk stomatal resistance, and critical root density (which is used to determine effective root-zone depth). The test is made in stand-alone mode, that is, using prescribed radiative and atmospheric forcing. Model performance is evaluated by comparing modeled runoff ratios with observed runoff ratios for a set of basins where precipitation biases have been shown to be minimal.

The withholding of information on global variations in these parameters leads to a significant degradation of the capability of the model to simulate the annual runoff ratio. An additional set of optimization experiments, in which the parameters are examined individually, reveals that the stomatal resistance is, by far, the parameter among these three whose spatial variations add the most predictive power to the model in stand-alone mode. Further single-parameter experiments with surface roughness length, available water capacity, thermal conductivity, and thermal diffusivity show very little sensitivity to estimated global variations in these parameters. Finally, it is found that even the constant-parameter model performance exceeds that of the Budyko and generalized Turc–Pike water-balance equations, suggesting that the model benefits also from information on the geographic variability of the temporal structure of forcing.

Full access
P. C. D. Milly and A. B. Shmakin

Abstract

A simple model of large-scale land (continental) water and energy balances is presented. The model is an extension of an earlier scheme with a record of successful application in climate modeling. The most important changes from the original model include 1) introduction of non-water-stressed stomatal control of transpiration, in order to correct a tendency toward excessive evaporation; 2) conversion from globally constant parameters (with the exception of vegetation-dependent snow-free surface albedo) to more complete vegetation and soil dependence of all parameters, in order to provide more realistic representation of geographic variations in water and energy balances and to enable model-based investigations of land-cover change; 3) introduction of soil sensible heat storage and transport, in order to move toward realistic diurnal-cycle modeling; 4) a groundwater (saturated-zone) storage reservoir, in order to provide more realistic temporal variability of runoff; and 5) a rudimentary runoff-routing scheme for delivery of runoff to the ocean, in order to provide realistic freshwater forcing of the ocean general circulation model component of a global climate model. The new model is tested with forcing from the International Satellite Land Surface Climatology Project Initiative I global dataset and a recently produced observation-based water-balance dataset for major river basins of the world. Model performance is evaluated by comparing computed and observed runoff ratios from many major river basins of the world. Special attention is given to distinguishing between two components of the apparent runoff ratio error: the part due to intrinsic model error and the part due to errors in the assumed precipitation forcing. The pattern of discrepancies between modeled and observed runoff ratios is consistent with results from a companion study of precipitation estimation errors. The new model is tuned by adjustment of a globally constant scale factor for non-water-stressed stomatal resistance. After tuning, significant overestimation of runoff is found in environments where an overall arid climate includes a brief but intense wet season. It is shown that this error may be explained by the neglect of upward soil water diffusion from below the root zone during the dry season. With the exception of such basins, and in the absence of precipitation errors, it is estimated that annual runoff ratios simulated by the model would have a root-mean-square error of about 0.05. The new model matches observations better than its predecessor, which has a negative runoff bias and greater scatter.

Full access
P. C. D. Milly and Krista A. Dunne

Abstract

Hydrologic models often are applied to adjust projections of hydroclimatic change that come from climate models. Such adjustment includes climate-bias correction, spatial refinement (“downscaling”), and consideration of the roles of hydrologic processes that were neglected in the climate model. Described herein is a quantitative analysis of the effects of hydrologic adjustment on the projections of runoff change associated with projected twenty-first-century climate change. In a case study including three climate models and 10 river basins in the contiguous United States, the authors find that relative (i.e., fractional or percentage) runoff change computed with hydrologic adjustment more often than not was less positive (or, equivalently, more negative) than what was projected by the climate models. The dominant contributor to this decrease in runoff was a ubiquitous change in runoff (median −11%) caused by the hydrologic model’s apparent amplification of the climate-model-implied growth in potential evapotranspiration. Analysis suggests that the hydrologic model, on the basis of the empirical, temperature-based modified Jensen–Haise formula, calculates a change in potential evapotranspiration that is typically 3 times the change implied by the climate models, which explicitly track surface energy budgets. In comparison with the amplification of potential evapotranspiration, central tendencies of other contributions from hydrologic adjustment (spatial refinement, climate-bias adjustment, and process refinement) were relatively small. The authors’ findings highlight the need for caution when projecting changes in potential evapotranspiration for use in hydrologic models or drought indices to evaluate climate-change impacts on water.

Full access
A. B. Shmakin, P. C. D. Milly, and K. A. Dunne

Abstract

The Land Dynamics (LaD) model is tested by comparison with observations of interannual variations in discharge from 44 large river basins for which relatively accurate time series of monthly precipitation (a primary model input) have recently been computed. When results are pooled across all basins, the model explains 67% of the interannual variance of annual runoff ratio anomalies (i.e., anomalies of annual discharge volume, normalized by long-term mean precipitation volume). The new estimates of basin precipitation appear to offer an improvement over those from a state-of-the-art analysis of global precipitation (the Climate Prediction Center Merged Analysis of Precipitation, CMAP), judging from comparisons of parallel model runs and of analyses of precipitation–discharge correlations. When the new precipitation estimates are used, the performance of the LaD model is comparable to, but not significantly better than, that of a simple, semiempirical water-balance relation that uses only annual totals of surface net radiation and precipitation. This implies that the LaD simulations of interannual runoff variability do not benefit substantially from information on geographical variability of land parameters or seasonal structure of interannual variability of precipitation.

The aforementioned analyses necessitated the development of a method for downscaling of long-term monthly precipitation data to the relatively short timescales necessary for running the model. The method merges the long-term data with a reference dataset of 1-yr duration, having high temporal resolution. The success of the method, for the model and data considered here, was demonstrated in a series of model–model comparisons and in the comparisons of modeled and observed interannual variations of basin discharge.

Full access
Kirsten L. Findell, Thomas R. Knutson, and P. C. D. Milly

Abstract

The Geophysical Fluid Dynamics Laboratory atmosphere–land model version 2 (AM2/LM2) coupled to a 50-m-thick slab ocean model has been used to investigate remote responses to tropical deforestation. Magnitudes and significance of differences between a control run and a deforested run are assessed through comparisons of 50-yr time series, accounting for autocorrelation and field significance. Complete conversion of the broadleaf evergreen forests of South America, central Africa, and the islands of Oceania to grasslands leads to highly significant local responses. In addition, a broad but mild warming is seen throughout the tropical troposphere (<0.2°C between 700 and 150 mb), significant in northern spring and summer. However, the simulation results show very little statistically significant response beyond the Tropics. There are no significant differences in any hydroclimatic variables (e.g., precipitation, soil moisture, evaporation) in either the northern or the southern extratropics. Small but statistically significant local differences in some geopotential height and wind fields are present in the southeastern Pacific Ocean. Use of the same statistical tests on two 50-yr segments of the control run show that the small but significant extratropical differences between the deforested run and the control run are similar in magnitude and area to the differences between nonoverlapping segments of the control run. These simulations suggest that extratropical responses to complete tropical deforestation are unlikely to be distinguishable from natural climate variability.

Full access
Jonghun Kam, Thomas R. Knutson, and P. C. D. Milly

Abstract

Over regions where snowmelt runoff substantially contributes to winter–spring streamflows, warming can accelerate snowmelt and reduce dry-season streamflows. However, conclusive detection of changes and attribution to anthropogenic forcing is hindered by the brevity of observational records, model uncertainty, and uncertainty concerning internal variability. In this study, the detection/attribution of changes in midlatitude North American winter–spring streamflow timing is examined using nine global climate models under multiple forcing scenarios. Robustness across models, start/end dates for trends, and assumptions about internal variability are evaluated. Marginal evidence for an emerging detectable anthropogenic influence (according to four or five of nine models) is found in the north-central United States, where winter–spring streamflows have been starting earlier. Weaker indications of detectable anthropogenic influence (three of nine models) are found in the mountainous western United States/southwestern Canada and in the extreme northeastern United States/Canadian Maritimes. In the former region, a recent shift toward later streamflows has rendered the full-record trend toward earlier streamflows only marginally significant, with possible implications for previously published climate change detection findings for streamflow timing in this region. In the latter region, no forced model shows as large a shift toward earlier streamflow timing as the detectable observed shift. In other (including warm, snow free) regions, observed trends are typically not detectable, although in the U.S. central plains we find detectable delays in streamflow, which are inconsistent with forced model experiments.

Full access