H31E-01 INVITED
The role of antecedent soil moisture on variability of the North American Monsoon System
We evaluate the influence of soil moisture anomalies on the timing and strength of North American Monsoon system (NAMS) precipitation through analysis of retrospective data sets including off-line simulations with the Variable Infiltration Capacity (VIC) land surface model, and through coupled model simulations using the MM5 mesoscale climate model coupled with the VIC land surface scheme. The role of land surface conditions on variations in monsoon precipitation in the Arizona-New Mexico and northwestern Mexico subregions of the North American Monsoon region are evaluated. The retrospective data analysis shows that soil moisture memory can propagate winter precipitation anomalies, and hence land surface cooling, through the dry spring season and into early summer. The effect is greater in NW Mexico where the monsoon begins earlier than in the southwestern U.S. We further investigate this land surface feedback mechanism through a set of coupled model runs using MM5/VIC. These coupled runs are consistent with the previous off-line runs to the extent that the VIC land surface scheme is the basis for soil moisture prediction in both. MM5/VIC control runs together with a set of sensitivity experiments in which soil moisture is prescribed to field capacity, wilting point and VIC soil moisture climatology, respectively, during pre-monsoon seasons (April-June) are used to examine the influence of antecedent (above-normal, below-normal and normal) soil moisture on pre-monsoon (May and June) surface temperature. Surface temperature, and its contrast with sea surface temperature, is a key driver of the onset of the NAMS. These experiments are intended to better understand the role of land-atmosphere feedbacks on the NAMS by testing a range of land surface and climate conditions in the coupled modeling environment.
H31E-02 INVITED
Understanding the Rainfall Daily Climatology of Northwestern Mexico
Maximum monthly precipitation (MMP) over northwestern Mexico is not concurrent because it occurs in different months from July through September. However, instead of occurring progressively from one month to the next as latitude increases, as it might be logic since rains move progressively from south to north as monsoon develops, MMP occurs in July in latitudes of Jalisco state, then MMP shifts to August more to the north in latitudes of Nayarit state and along the eastern coast of the Gulf of California, then it occurs in July in higher latitudes through the main axis of the Sierra Madre Occidental (SMO), and finally MMP shifts to September to the west in the California Peninsula. The maximum monthly streamflow occurs in a similar pattern as MMP does but one month later. When daily rainfall climatology of the region is calculated, i.e. the long-term mean per day from stations with more than 20 years of data between 1940 and 2004, it is possible to understand why the behavior of MMP occurs in a July-August-July pattern from south to north. Preliminary results indicate that at latitudes of Nayarit state normal frequent storms with abundant rains develop at the end of July and through the August. These rains sum to the rains that move from the south to the north, as monsoon develops increasing the volume of precipitations at those latitudes in August. To the east crossing the SMO through northwestern Zacatecas state maximum volume of precipitations also is observed in August. However, in higher latitudes it is not observed any increment of rains in August and consequently maximum volume of precipitations occurs in July. To understand the dynamics of the rains at the latitudes of Nayarit state it results necessary to investigate the source of these local rains and explain why the increase of precipitations in August is limited at those latitudes.
H31E-03 INVITED
Precipitation Recycling: a Mechanism for Hydroclimatological Stability in the North American Monsoon Region
In this work we study precipitation recycling as part of the dynamic North American Monsoon System (NAMS), using a set of land-atmosphere variables derived from North American Regional Recycling (NARR) data. The recycling ratio is estimated using the Dynamic Recycling Model which provides recycling estimates at daily timescales. We show that precipitation recycling is a significant source of moisture for monsoon precipitation, with recycling ratios consistently above 15% and sometimes as high as 25%. The 1985-1995 climatological analysis of NAMS precipitation recycling reveals a positive feedback mechanism between monsoon precipitation and subsequent increase in precipitation of recycled origin. One of the innovative features of this analysis is the delineation of the source and sink regions of recycled precipitation within the NAMS domain. While monsoon precipitation and evapotranspiration are predominantly located in the seasonally dry tropical forests in the southwestern part of the domain, recycling is enhanced northeast of this region, indicating a relocation of soil moisture further inland to drier regions in the northeast. Interestingly, the three years with longest monsoons in the 11-year period present an a-synchronous pattern between precipitation and recycling ratio. The longest monsoons have a characteristic double peak in precipitation, with enhanced recycling ratios during the intermediate dry period. Indicating that, even when large scale moisture advection decreases, evapotranspiration keeps providing moisture to the overlying atmosphere and contributing to precipitation. Through the negative feedback present during long monsoons and by relocation of soil moisture, precipitation recycling brings favorable conditions for vegetation sustenance in the NAMS region.
H31E-04
On the disaggregation of satellite based passive microwave estimates of soil moisture: current status and future challenges
Soil moisture is a fundamental state variable that controls several earth surface related processes, i.e. hydrology,
meteorology, climate modelling, and agricultural management. However, the spatial and temporal dynamic of soil
moisture dynamic is very complex since it depends on several factors such as weather condition, land cover/land
use, soil type, topography, geology. Capturing such dynamic requires a dense network of continuous observation
of soil moisture which is not feasible. The only realistic possibility for derive continuous spatially distributed soil
moisture is through satellite observations. In this regard Passive microwave sensors, especially those operating
at low frequencies (L bands) present an interesting potential for monitoring soil moisture. However, the use of
coarse spatial resolution of instrument such as SMOS in the field of hydrology is not straightforward. Indeed, the
scale at which most hydrological processes occur is approximately 1km or less. It is thus of crucial importance to
develop procedures to disaggregate passive microwave based soil moisture from its nominal scale to that
needed for hydrological application and/or watershed management. The objective of this presentation is to
provide an overview of existing and newly developed techniques for disaggregating soil moisture from coarse
scale to scale relevant for hydrological application. Ground and aircraft data collected at the Walnut Gulch
experimental watershed are used to discuss the performance and the limitation of these approaches.
http:www.cesbio.ups-tlse.fr
H31E-05
Hydroclimatic variability in the North American Monsoon region of northwest Mexico
River systems in southwestern North America exhibit a strong gradient in their seasonal runoff climatologies which are attributed to regional transitions in climatic forcing. Those basins draining the Sierra Madre Occidental of western Mexico are strongly influenced by summer monsoon rainfall while river systems in the southern and central Rocky Mountains are largely snowmelt or cool-season dominated. While much research has focused on the interannual variability of streamflow in snow-melt dominated basins in the U.S., comparatively less research has documented the interannual variability and long-term changes in summer streamflow in northwest Mexico. In this work, interannual variations in regionalized streamflow and rainfall-runoff relationships from northwest Mexico are explored. Long term trends between summer and cool-season streamflow exhibit markedly different structures in some regions. Partitioning between cool season and warm season streamflow shows a marked transition in correspondence with the 1976-1977 climate regime shift as well as significant interannual modulation that appears to be linked to ENSO. Similarly, significant changes in the fraction of seasonal streamflow to seasonal rainfall (ie runoff fractions) exhibit an ENSO-like influence. Recent work on characterizing the relationship between seasonal precipitation structures across the SMO and runoff responses during both the warm and cool seasons will be presented.
H31E-06
Multi-scale field investigation of water flow pathways and residence times in mountainous catchments during monsoon rainfall
The "sky islands" of Arizona and New Mexico in the southwestern United States form a unique complex of about 27 mountain ranges whose ecosystems support many perennial and ephemeral streams in an arid climate. Among these sky islands are the Santa Catalina Mountains near Tucson, AZ, with a peak elevation of 9157 ft at Mt. Lemmon. Sabino Canyon Creek is the main stream which runs on the south face of the mountain range. It usually flows from July through April with an average daily flow of approximately 0.28 m3/s (10 cfs). However, flash floods are common both during summer as a result of intense monsoon rains and during spring because of rapid snowmelt. During these events, flow increases drastically, reaching peak flows up to 480 m3/s (15,984 cfs, July 2006). Characterizing water flow pathways and residence times in these complex catchments is important for improving flash flood warning systems, estimating mountain front recharge, managing forest and wild fires, and understanding ecosystem functions. In the summer of 2006, we set up an extensive hydrometrical and hydro- chemical monitoring network in Sabino Canyon Creek, comprising 40 tipping bucket rain gauges (two of which were equipped to automatically collect rainwater samples), 5 automatic surface water level stations (three of which were equipped with auto samplers), and 8 manual soil lysimeters. In addition, several rain and stream water grab samples were collected manually during intensive rain events. Water samples are analyzed for major ions and liquid water isotopic concentration (2H and 18O) in rain, soil, ground and surface water. The data allows for a detailed reconstruction of water flow pathways and residence times at 3 different catchment scales (2 km2, 8 km2, and 91 km2) during the recorded flow events, including the highest monsoon rainfall-runoff event ever recorded in these mountains.