Specific Priorities
Short-term convective weather forecasting
Accurate short-period (0-6h) precipitation forecasts would benefit many sectors of society including aviation, water management, flood forecasting, emergency management, recreation, and building industries. Current techniques do not provide reliable convection forecasts that meet these user needs. The skill of observation-based nowcasts decreases rapidly with lead time. NWP has limited predictive ability in the first few hours due primarily to the problem of model spin-up. The purpose of the research and development efforts described here is to bridge the gap in skill represented by these forecast techniques through the optimal combination of observation based nowcasts and numerical modeling.
A key effort for the next five years is to focus and organize research in this area across NCAR through the Short-Term Explicit Prediction (STEP) program to address seven main topic areas required to improve short period forecast: basic studies of the predictability; observations (data acquisition, quality control and data analysis); data assimilation; numerical weather prediction; nowcasting; physical parameterization; and verification and end-user needs.
Four-dimensional Variational Data Assimilation (4DVar)
An advanced four-dimensional variational data assimilation system (4DVar) is being used to assimilate radar observations of the atmosphere into a numerical model. The 4DVar approach seeks to find the initial conditions of the model that minimizes the fit to the observations (primarily radar) and a background field. Tests of the system in research mode over a numbers of years have shown that it has the ability to provide accurate short-term forecasts of convective storms. The system has been ported to the operational environment where it runs in real time in both Indiana/Illinois and Dallas/Fort Worth and assimilates data from four to five WSR-88D radars. Research activities will be conducted to transfer the four dimensional variation technique to the WRF model and apply this methodology to the 0-6 h forecast problem. RAL will also continue to conduct real-time demonstrations of the system as well as test and evaluate the system for winter weather forecasting. Improvements to water vapor and temperature assimilation will also be made.
Observations of near-surface moisture
To make substantial improvements in precipitation forecasts, better measures of atmospheric moisture are required. The REFRACTT (Refractivity Experiment For H2O Research And Collaborative operational Technology Transfer) program was designed and conducted to demonstrate a methodology, developed at McGill University, to measure near-surface atmospheric moisture using radar data. The refractivity field describes the near-surface index of refraction using radar phase measurements from ground targets. Water vapor estimates are derived based on the refractivity metrics and temperature from nearby mesonet stations. The water vapor measurements from the field program are also being assimilated into the VDRAS model to determine potential improvements in very short period explicit model forecasts.
Composite radar reflectivity (left) and refractivity (right) from the Denver WSR-88D, SPOL and CHILL radars. A line of storms has moved in from the west and an associated radar reflectivity fine line marks the leading edge gust front, which is labeled on both images.
RAL will participate in a second REFRACTT field program in the Colorado Front Range. Modeling studies will be continued to determine how to optimally use the REFRACTT water vapor data to improve convective weather forecasts. Staff will also work with EOL and university researchers to investigate the use of dual-frequency radar data to estimate water vapor using the NSF S-PolKa (S and Ka-band) radar.
International aerosol and precipitation studies
Over the past few years, RAL research has focused on analyzing existing data and collecting specialized data for assessing current weather modification projects or for developing feasibility plans for cloud seeding activities in the UAE, India, Saudi Arabia, Oman, Indonesia, and Italy. Training of local scientists and engineers in scientific research and field operations has been an important facet in each of these programs. In the future, RAL will conduct multi-parameter radar-based field programs to study aerosol-precipitation interaction in convective and orographic clouds and participate in a SE Asia regional study on aerosols and precipitation in collaboration with NASA and other agencies. In-situ validation of satellite-based aerosol measurements, especially over land, are needed to establish the utility of this information. Working with local African agencies, RAL will also collaborate with the African Monsoon Multidisciplinary Analyses (AMMA) field program on aerosol-precipitation interactions and the characterization of convective cloud organization. C-band radars are relatively easy to deploy for these studies; RAL will test and evaluate advanced multi-parameter hydrometeor identification techniques for these radars.
Winter snowpack enhancement
The Wyoming Water Development Commission, on behalf of the State of Wyoming, is overseeing the Wyoming Weather Modification Pilot Project, a winter cloud seeding program to increase snowpack and runoff within Wyoming’s Green River, Wind River, and Platte River basins. This program involves cloud seeding operations (conducted by Weather Modification, Incorporated) and scientific evaluations by RAL scientists. Two different geographic areas will be subjected to cloud seeding in an attempt to increase precipitation (snowpack) over targeted mountainous areas by 10-20% per year – amounts within the range of natural variability between winter seasons and for individual storms. RAL will participate in the pilot projects during which clouds will be seeded from airborne and ground-based generators and will analyze data from the first season to refine the experimental design for future field efforts. The ability of the WRF model to simulate the formation of super-cooled liquid water over complex terrain will be investigated using the Wyoming in-situ data as verification. These studies will lead to a determination of the potential for silver iodide cloud seeding to increase snowpack over the complex terrain in Wyoming.
Romanian Destructive Waters Abatement Program (DESWAT)
RAL scientists are currently providing modeling support for a World Bank-funded project for Romania. Under subcontract from Baron Advanced Meteorological Services (BAMS) and in collaboration with the Land Information Team from NASA Goddard Space Flight Center (GSFC), a series of hydrological enhancements to the Noah land surface parameterizations are being developed and implemented which will enable runoff and stream flow predictions. Once completed this new modeling system will provide a real-time water resources assessment and forecasting tool for Romania in support of its application to enter the European Union. RAL scientists are adapting proven watershed modeling techniques for use within the Noah model, which was originally developed for land surface modeling within weather and climate models. The approach used employs physically-based, dynamical routing techniques of surface and subsurface water across the landscape and through both natural and managed river systems. In addition to developing a hydrologic prediction system which is robust to variety of implementations, the modifications are also intended to be modularized and extensible for easy coupling to the NCAR supported Weather Research and Forecasting (WRF) model. In the future, RAL will work to couple Noah-distributed with an automated parameter estimation model to add in calibration and optimization and perform extensive operational and offline testing of the model both within Romania and in collaborative inter-comparison projects (e.g. DMIP-II) in the U.S. The Noah-distributed model will be adapted for customizable, multi-scale deployment within the NCAR High Resolution Land Data Assimilation System.
North American Monsoon Experiment (NAME)
RAL staff implemented and operated a network of 87 tipping bucket rain gauges in the remote regions of the Sierra Madre Occidental Mountains in northwest Mexico; the locus of the North American Monsoon System. A major activity of this project is the regular processing and quality control of the data from the network. RAL is collaborating with the University of Arizona on the development of a statistical downscaling model which is to be used in support of seasonal hydrologic forecasts for the NAME region. Once completed the suite of forecasting tools generated under this project will provide a robust, community-model based platform from which seasonal forecasts of precipitation and stream flow will be made. RAL priorities for the future include diagnostic analyses of NAME precipitation, and assessment of the influence of teleconnection-type forcing mechanisms on monsoon hydrology. RAL staff will participate in the NAMAP2 (North American Monsoon Model Assessment Program Phase 2), and continue operation and maintenance of the NAME Event Rain gauge Network in northwest Mexico. A statistical downscaling model for seasonal hydrological forecasts to other regions will be adapted for use in this project. Short-term and intra-seasonal forecasts of monsoon stream flow will be compared using this statistical method versus the physically-based Noah-distributed model.
Implementation of a subgrid land surface modeling framework within the CLM
Through support from the NSF Biocomplexity program, RAL developed and implemented a framework for enabling sub-grid execution of the Common Land Model (CLM) within the NCAR CCSM framework. Implementing such a framework allows scientists to better engage terrestrial processes and the landscape scale and conduct detailed investigations on multi-scale processes such as terrestrial hydrology, land-use changes, water and carbon cycle linkages in a manner heretofore inhibited. The project is now nearing completion of its first phase and a prototype version of the enhanced CLM model is now available. RAL will continue this work by conducting numerous climate scenario simulations to assess the impact of the sub-gridding and downscaling techniques on climates using the CCSM and CAM modeling systems. RAL will also engage the CLM community in the collaborative development of downscaling and up-scaling algorithms for use in the new CLM, continue the development of scale-appropriate physical parameterizations for use within the sub-grid CLM model, and migrate the new sub-grid version of CLM to the WRF regional climate model
California water resources
RAL collaborated with the Stockholm Environment Institute-Boston (SEIB), the Natural Heritage Institute (NHI) of Sacramento California, and faculty from the University of California, Davis and Berkley on a study of freshwater ecosystem services in the Sacramento Basin of Northern California. The study team has developed an integrated water resource modeling framework that can be used to investigate medium to long term water resource planning and management issues throughout the Sacramento Basin. This framework makes use of climate data derived from a new Bayesian analysis technique that yields statistical distributions of regional climate change based on regional projections from multiple Atmosphere-Ocean models. RAL will seek additional partners to continue funding for this type of research and adapt the framework for climate change impacts on water use, demand, and hydropower generation
Emerging product development
In conjunction with the DESWAT project described, RAL is in the process of releasing an open source, parallelized, offline hydrological prediction system called Noah-distributed. This tool will serve as a framework from which further interaction with the hydrological community will occur.
Future priorities are to couple the Noah-distributed model to the community WRF model to create the WRF-Hydro community regional water cycle model. This new modeling framework will provide a framework from which to examine both terrestrial and atmospheric hydrological processes across a range of temporal and spatial scales and will form the backbone of the WRF-based regional climate model.
ALERT
Denver’s Urban Drainage and Flood Control District (UDFCD) manage the ALERT (Automated Local Evaluation in Real Time) urban flood warning system whose aim is to reduce injuries, deaths, and property damage caused by floods. RAL’s work on ALERT has been to demonstrate the usefulness ofNCAR’s Thunderstorm Identification Tracking and Analysis (TITAN) data in support of the UDFCD flood threat monitoring and dissemination mission. This work includes development of basin forecast products based on radar analysis, construction of rainfall-runoff model for some watersheds and delineation of specific basins for additional development. Additional development, testing, and evaluation are needed to gain wider acceptance of the product, including improved precipitation estimates and forecasts, and extension of the system’s analytical capability to encompass all important UDFCD watersheds.
Diagram of the ALERT flood warning system employing RAL’s TITAN software
Microphysical parameterizations
Research activities related to microphysical modeling are designed to improve the simulation of cloud water (including super-cooled cloud water), rain and drizzle, freezing rain and freezing drizzle, snow, snow pellets, ice pellets and hail in RUC and WRF models. Super-cooled cloud water and freezing drizzle are emphasized due to their importance to aircraft icing. This research also is designed to improve forecasts of winter weather conditions at the ground, as well as improving Quantitative Precipitation Forecasting (QPF). RAL will lead the planned Ice Initiation Field Study to take place winter 2006-2007 in northeastern Colorado (NSF proposal and spring 2006 OFAP request) and will seek additional opportunities to conduct field studies on a “piggyback” basis. Model upgrades will focus on adding improved treatments of cloud-active aerosol (ice nuclei and CCN) advection, sources, and sinks to MM-5. In addition, more accurate simulation of hydrometeor melting and evaporation are being sought in order to predict more accurately the formation of downdrafts. The performance of the improved microphysical model will be compared with observations. As part of technology transfer, new microphysics parameterizations will be implemented and tested in WRF and its various special versions.