Application of Next-Generation Satellite Data to a High-Resolution, Real-Time Land Surface Model

Zavodsky1Applied Sciences Theme, Articles, Earth Observation, Original, Technology

Table showing the schematic of a single SPoRT-LIS cycle. Each cycle is run every six hours from five days (d) before the current time (t) to nine hours (h) past the current time to provide the most up-to-date soil moisture information for situational awareness and model initialization.

Bradley T. Zavodsky1, Jonathan L. Case2, Clay B. Blankenship3, William L. Crosson3, Kristopher D. White4
1NASA/Marshall Space Flight Center, Huntsville, Alabama
2ENSCO, Inc., Huntsville
3Universities Space Research Association, Huntsville
4NOAA/National Weather Service, Huntsville

Figure showing a schematic of a single SPoRT-LIS cycle.  Each cycle is run every six hours from five days (d) before the current time (t) to nine hours (h) past the current time to provide the most up-to-date soil moisture information for situational awareness and model initialization.

Figure 1: Schematic of a single SPoRT-LIS cycle. Each cycle is run every six hours from five days (d) before the current time (t) to nine hours (h) past the current time to provide the most up-to-date soil moisture information for situational awareness and model initialization.

Operational application of NASA satellite observations is identified as a key component of the Weather Focus Area within the NASA Earth Science Division. Thus, next generation NASA instruments, such as the Soil Moisture Active Passive (SMAP) and Global Precipitation Measurement (GPM) missions have an inherent operational objective, and the potential for the data to improve operational forecasts must be explored. This paper describes activities at the Short-term Prediction Research and Transition (SPoRT) Center to develop a suite of high-resolution, near-real-time land surface model (LSM) products integrating these two datasets that will meet the requirements of National Oceanic and Atmospheric Administration (NOAA) and National Weather Service (NWS) local operational forecasting entities. This product suite will use the current operational SPoRT-Land Information System (LIS) platform to assimilate enhanced soil moisture estimates from SMAP L-band observations and incorporate gridded precipitation analysis products from the GPM constellation of satellites to generate improved LSM results. It is anticipated that improvements to both numerical weather prediction and situational awareness for forecast challenges such as drought monitoring, excessive heat during dry-soil conditions, and convective precipitation will result from exploitation of data products from these missions.
I. Introduction
Proper characterization of soil moisture on local spatial scales plays a large role in how well meteorological features can be forecasted by operational meteorologists, like those at the National Oceanic and Atmospheric Administration (NOAA) and National Weather Service (NWS). Next generation satellites—such as the NASA Soil Moisture Active Passive (SMAP; [1]) mission and the international constellation of satellites that make up the Global Precipitation Measurement (GPM; [2]) mission led by the NASA/Japan Aerospace Exploration Agency (JAXA) GPM Core satellite—will provide advanced datasets to improve soil moisture characterization for a number of operational weather forecasting challenges. While NASA instruments have traditionally been viewed as research satellites, these next-generation satellites have an inherent applications objective as NASA partners with other agencies to exploit the greatest utility of the data to benefit the general public through application to operational weather forecasting (e.g., [3], [4]).
To impact local-scale operations, products must be able to capture fine-scale features and be timely enough to meet the stringent deadlines on issuing forecasts to the general public. The initialization of land surface variables in local models typically involves interpolating grids from large-scale models, which may not provide representative fields consistent with the local model state or grid resolution [5]. Assimilation of soil moisture observations from SMAP and incorporation of global precipitation analyses from GPM have the potential to improve land surface model (LSM) initializations. The current products planned for the SMAP and GPM missions will not meet one or both of these operational requirements as stand-alone products, so direct transition may not be effective for the local forecasting process. Additional products will need to be developed for the local forecaster.
Improving local-scale operational weather forecasts through use of satellite data is the key objective of the Short-term Prediction Research and Transition (SPoRT; [6]) Center at the NASA Marshall Space Flight Center. SPoRT works collaboratively with operational forecasters at the NWS to transition new satellite data sets to local forecasters through development and assessment of unique products that meet their needs. To address the limitations of LSMs for local forecasters, SPoRT currently runs the NASA Land Information System (LIS; [7], [8], [9]) in real time, which provides soil characteristics for situational awareness applications and incorporation into local numerical weather prediction (NWP) models.
What follows is an overview of the current configuration of the real-time SPoRT-LIS, operational applications of the high-resolution SPoRT-LIS output, and the expected use and impact of SMAP and GPM data to enhance the SPoRT-LIS product suite.
II. Operational SPoRT-LIS
SPoRT currently runs a real-time version of the NASA LIS that uses version 3.2 of the Noah LSM [10], [11] in an uncoupled (offline) mode. In this mode, the LSM runs independently of an NWP model with input variables provided either by large-scale atmospheric analyses or gridded observation datasets. The offline LIS/Noah currently runs over the Southeastern portion of the contiguous United States (CONUS), with plans to expand the domain to full CONUS. The 4-kilometer grid of the full CONUS SPoRT-LIS domain is of a high enough resolution to meet the requirements of local operational forecasters. Output grids for the SPoRT-LIS include volumetric soil moisture and soil temperature in the four operational Noah LSM layers (0-10, 10-40, 40-100 and 100-200 centimeters), snow-water equivalent, and skin temperature.
Currently, the SPoRT-LIS simulation consists of a long-term integration of the Noah LSM spanning multiple years forced by the North American Land Data Assimilation System Phase 2 (NLDAS-2; [12]) and the National Centers for Environmental Prediction (NCEP) Stage IV precipitation analysis product [13],[14]. For the real-time products, additional forcing data from the NCEP Global Data Assimilation System (GDAS; [15]) is required to drive the integration due to the latency of the NLDAS-2 product. To make the product available for real-time applications, short-term forecasts from the NCEP Global Forecast System (GFS; [16]) model are used in each of the six hourly SPoRT-LIS cycles in place of the GDAS analyses for future times. It is this real-time aspect of the SPoRT-LIS product that makes it most useful for local operational forecasters. Figure 1 is a schematic of the current real-time SPoRT-LIS methodology; more details can be found in [17].
Output grids from the SPoRT-LIS are currently used in two ways by regional and local operational weather forecasters. The first way is direct interpretation of the LSM grids for situational awareness related to drought monitoring, extreme heat, and convective initiation. The second way is initialization of high-resolution, small-domain local NWP simulations performed by numerous NWS Weather Forecast Offices (WFOs) to provide high-resolution guidance for predicting these and other phenomena. The SPoRT-LIS output for situational awareness applications is available to forecasters in real-time via a local data manager. For NWP applications, SPoRT-LIS output is posted in real time to an anonymous file transfer protocol (FTP) system from which the data are acquired automatically by the NOAA/NWS Science and Training Resource Center (STRC) Environmental Model System (STRC-EMS or EMS; [18], [19]). The EMS is a packaged version of the Weather Research and Forecasting (WRF; [20]) model that incorporates nearly every element of an operational NWP system, including acquisition and processing of initialization data, model execution, output data processing, and file management [18]. Individual NWS WFOs uniquely configure and run the EMS on local workstations to address specific forecast challenges. The official EMS software provides an input option for using the SPoRT-LIS dataset in place of large-scale, coarser-resolution model input.

Figure of the United States Drought Monitor (USDM) before (a) and after (c) a local drought event.  The SPoRT-LIS soil moisture product (b) was used to extend the Moderate D1 drought category southwestward in Northeastern Alabama (circled area).  Adapted from {22}.

Figure 2: United States Drought Monitor (USDM) before (a) and after (c) a local drought event. The SPoRT-LIS soil moisture product (b) was used to extend the Moderate D1 drought category southwestward in Northeastern Alabama (circled area). Adapted from {22}.

III. SPoRT-LIS Applications
A. Drought Monitoring
Drought monitoring is one application of the SPoRT-LIS for operational forecasters [21], [22]. By definition, drought is a deficit of moisture of sufficient temporal and spatial magnitude to cause social, environmental, or economic impacts. The quantification of drought conditions involves the synthesis of precipitation, hydrologic, soil moisture, agricultural, and ecological datasets. Historically, soil moisture has been one of the most difficult parameters to determine due to the sparse spatial coverage of in-situ networks. While regional drought conditions are typically detected in large-scale analysis products, there are times where these analyses cannot resolve fine-scale dry features that arise from heterogeneous rainfall patterns, which can be particularly prevalent during the growing season. The inability to resolve drought on local scales may result in incorrect decisions regarding federal drought aid for farmers. White and Case [22] describe how the Applications Integration Meteorologist at the NWS office in Huntsville, Alabama, uses real-time SPoRT-LIS output to assess soil moisture deficits and surpluses and relays the sub-county scale information to decision makers at the U.S. Drought Monitor (USDM). An example of how the SPoRT-LIS output has been used to make adjustments to the USDM is shown in Figure 2. In this example, the higher-resolution SPoRT-LIS enabled decision-makers to extend the moderate drought conditions in Northeast Alabama southwestward. Assimilation of higher-resolution soil moisture observations into the SPoRT-LIS should enable officials to more accurately pinpoint local counties most heavily affected by drought.

Figure showing anomalies of (a) soil moisture (mm of water) in upper meter on June 30, 2012, and (b) mean daily maximum temperature (in degrees Fahrenheit) for seven days, ending July 3, 2012.

Figure 3: Anomalies of (a) soil moisture (mm of water) in upper meter on June 30, 2012, and (b) mean daily maximum temperature (in degrees Fahrenheit) for seven days, ending July 3, 2012.

B. Maximum Heat in Dry Soil Conditions
Extreme heat is a serious public health issue, with heat waves causing more deaths annually in the U.S. than any other extreme weather events [23], [24]. The NWS employs a heat warning system based on forecast temperature and heat index, which combines the effect on humans of temperature and humidity. During extreme heat episodes, small differences in temperature can drastically change the impacts on human health, so accurate forecasts of air temperature and humidity are critical for effective public warnings. Exceptionally high temperatures frequently occur when the soil is anomalously dry, since under these conditions sensible heat flux dominates the surface energy balance (Figure 3). NWP models tend to underestimate daytime air temperatures when the modeled soil moisture is higher than the true conditions (Figure 4), which may limit the effectiveness of public heat warnings. Improved specification of surface boundary conditions has been shown to lead to more accurate surface energy fluxes and boundary layer temperatures [25], [26]. More accurate, higher-resolution soil moisture observations from the SPoRT-LIS should result in improvements to these extreme maximum temperature forecasts.

Figure showing Model Output Statistics (MOS) from NAM forecasts (red) and observed (black) daily maximum temperatures for Huntsville, Alabama, during an extreme heat event during the summer of 2012.  Red squares represent the mean maximum temperature of the NAM forecasts ranging from 0-48 hours and the red bars represent the range of those NAM forecasts.

Figure 4: Model Output Statistics (MOS) from NAM forecasts (red) and observed (black) daily maximum temperatures for Huntsville, Alabama, during an extreme heat event during the summer of 2012. Red squares represent the mean maximum temperature of the NAM forecasts ranging from 0-48 hours and the red bars represent the range of those NAM forecasts.

C. Convective Initiation
Convective initiation under weakly-forced atmospheric regimes during the summer is one of the most challenging weather forecast problems. Given sufficient instability, deep moist convection often initiates in response to local low-level convergent boundaries that arise from horizontal gradients in surface heating rates, which convection-allowing NWP models can have trouble representing accurately. Under such regimes, a proper land surface initialization can be quite important for local models’ development of surface heating heterogeneity that can lead to convective initiation. Gradients in soil moisture (and subsequent surface heating rates) can lead to mesoscale circulations capable of affecting the development and intensity of clouds and precipitation systems (e.g., [27], [28], [29]). Additionally, local wet soil conditions can lead to a positive feedback mechanism on increased net radiation and low-level atmospheric water vapor [30], which in turn produces more unstable conditions and increases the likelihood of convection. The opposite feedback also can occur with dry soils [30]. Case et al. [5] showed that use of higher-resolution soil moisture designation has substantial impact on sensible heat flux, 2-m dew point, planetary boundary layer height, and convective available potential energy, all of which are important metrics for diagnosing the onset of convection in NWP models. Figure 5 shows differences between the 0-10 centimeter volumetric soil moisture initialization interpolated from a large-scale model versus that from a higher-resolution LIS configuration, demonstrating the small-scale details that are missed when land surface data from lower-resolution model fields are used to initialize local models. The soil moisture differences in Figure 5 translate into local changes in the simulated heat flux distribution (Figure 6a) and low-level moisture (Figure 6b), which has a corresponding effect on the evolution of the simulated boundary layer and convective instability (Figures 6c and 6d). Other studies also have shown promising improvements in convective initiation detection for local-scale NWP [31], [32]. It is anticipated that higher-resolution soil moisture observations from SMAP will capture many of these small-scale differences, which should result in improved initial conditions for NWP models.

Table showing the characteristics of currently planned Level 2 retrieved soil moisture products (from {39}, {40}, and {41}).

TABLE 1. Characteristics of currently planned Level 2 retrieved soil moisture products (from {39}, {40}, and {41}).

IV. Expected Significance of Data from Upcoming Missions
As described in Section II, the SPoRT-LIS is currently driven by GDAS atmospheric analyses, NLDAS-2 land information, and Stage IV precipitation. However, each of these datasets has a limitation to their application for local-scale modeling in the CONUS. Through integration of SMAP and GPM products into LIS to either enhance or replace the current analysis products, an enhanced SPoRT-LIS will be generated. It is hypothesized that these enhancements will result in even better representation of local-scale soil characteristics than the current product.
The modeled soil moisture fields could be improved in two ways from the SMAP and GPM missions. First, the assimilation of SMAP Level 2 soil moisture data can provide adjustments or corrections to the nominal LIS offline runs, especially in areas where the atmospheric forcing datasets have lower fidelity. Second, the gridded precipitation products from GPM could be used to supplement ground-based precipitation datasets in data-poor regions where radar and rain gauge coverage is sparse. GPM-derived precipitation fields could be particularly valuable for running real-time LIS applications in more remote regions of the globe where routine ground-based precipitation datasets are not readily available.

Figure showing Comparison between WRF-initialized 0-10 centimeter volumetric soil moisture for the a) control (12-kilometer, NAM-218), b) LIS spinup, and c) different field (LIS minus NAM) valid at 0300 UTC, June 9, 2008.  Adapted from {5}.

Figure 5: Comparison between WRF-initialized 0-10 centimeter volumetric soil moisture for the a) control (12-kilometer, NAM-218), b) LIS spinup, and c) different field (LIS minus NAM) valid at 0300 UTC, June 9, 2008. Adapted from {5}.

The NASA SMAP satellite comprises an L-band synthetic aperture radar coupled with a passive microwave radiometer. The L-band frequency allows for observations of soil moisture through moderate vegetation cover in clear and cloudy conditions and during both day and night. A number of Level 1 through 4 products are planned, but of specific interest for the SPoRT-LIS are three Level 2 retrieved soil moisture products (Table 1). In particular, the 3-kilometer spatial resolution of the radar soil moisture is expected to capture small-scale soil moisture features that cannot be resolved by current global products, making SMAP ideal for the 1-to-4 kilometer scales typical of local NWP.
SMAP soil moisture data will be assimilated into the Noah LSM within LIS using an Ensemble Kalman Filter (EnKF; [33], [34]) algorithm. An EnKF is a method to update an ensemble of model states (with a distribution reflecting the uncertainty in the model) with observations, producing a Bayesian optimal estimate of the new model state and its uncertainty. Here, the background (previous forecast) model state consists of soil moisture, temperature, and some surface variables; the observations are SMAP soil moisture. The new state depends on both the previous state and the observations, weighted by the relative uncertainty of the background and observations. A forward operator is required to convert the model fields into an observation variable. For the Level 2 soil moisture data, the forward operator will calculate 0-5 cm volumetric water content.
An enhanced SPoRT-LIS configuration will follow similarly to the current real-time SPoRT-LIS setup described in Section II and documented in [17]. In the enhanced setup, each LIS/Noah LSM re-start will begin a few days prior to the current time in order to assimilate the most recent SMAP observations using the LIS-EnKF and to incorporate the most recent GDAS/GFS forcing datasets. The LIS integration with SMAP data assimilation will continue through the time of the most recent SMAP data product (about 24 hours prior to the current time; see Table 1). At this time, the simulation will continue using additional forcing data from the GDAS/GFS short-term forecasts but will still retain information from the assimilated SMAP data. As a result of this “cycling” methodology, the impact of past SMAP observations is extended to the current time and enables the product to meet the data latency requirements of operational users, despite the expected 24-hour latency of the Level 2 products.

Figure showing Difference plots (LIS minus control) of a 13-hour forecast valid at 1600 UTC, June 9, 2008, for the following fields:  a) sensible heat flux (W m-2), b), 2-m dew point temperature (in degree Celsius), c) PBL height (m), and d) CAPE (J kg-1).  Adapted from {5}.

Figure 6: Difference plots (LIS minus control) of a 13-hour forecast valid at 1600 UTC, June 9, 2008, for the following fields: a) sensible heat flux (W m-2), b), 2-m dew point temperature (in degree Celsius), c) PBL height (m), and d) CAPE (J kg-1). Adapted from {5}.

The GPM mission consists of a core satellite housing a passive microwave rain radiometer called the GPM Microwave Imager (GMI) and dual-frequency precipitation radar (DPR) alongside an international constellation of satellites containing GMI and GMI-like rain radiometers. The concept behind the GPM mission is exploiting information from multiple current and future satellites along with ground based systems to obtain global precipitation for climate and weather studies. One of the expected products from GPM is the NASA GPM Integrated Multi-satellitE Retrievals for GPM (IMERG) Level 4 product, which will provide global, high-temporal (30 minutes) and high-spatial resolution (0.1 degree) analyses, which will blend gauge data with satellite-based radar, infrared, and passive microwave observations [35].
While Stage IV data provide coverage for much of the CONUS, there are areas where radar and gauge coverage is lacking, such as Southern Canada, Northern Mexico, and large areas of the Western U.S. [17]. Some of these data-void regions are upstream of areas of forecast responsibility to NWS forecasters (e.g. convective initiation over the Southwest and Southern U.S. may be related to air masses that originate in Mexico). Additionally, spurious model convection can result from artificial gradients that may appear in the soil fields along borders of the U.S. with Canada and Mexico as a result of using discontinuous precipitation data to force the LIS. Thus, satellite-based precipitation estimates, such as IMERG, will have the potential to enhance or replace Stage IV as the precipitation forcing data set within the enhanced SPoRT-LIS.
Incorporation of the GPM IMERG precipitation analyses is expected to be much more straightforward than the assimilation of the SMAP data. The GPM analysis will either simply replace the current Stage IV analyses or be used in data-void areas in the current Stage IV analysis as the precipitation forcing dataset used to drive the enhanced SPoRT-LIS integration.
C. Pre- and Post-launch Efforts
Prior to the launch of SMAP and GPM, legacy datasets are currently being used to determine optimal data use strategies so that the enhancements to the SPoRT-LIS will be available shortly after the launch of each mission. Pre-launch datasets are most valuable for understanding the characteristics and format of the data. The knowledge gained from the pre-launch datasets allows for the mechanics of integrating observations and products from new satellites to be performed prior to launch to enable more timely application of post-launch data. Soil moisture observations from the Soil Moisture Ocean Salinity (SMOS; [36]) instrument and simulated SMAP proxy data are being tested in a research version of the SPoRT-LIS to understand the potential impacts of the SMAP data on the LIS EnKF algorithm. Similarly, satellite precipitation products such as the Climate Prediction Center morphing method (CMORPH; [37]) and an experimental GOES Quantitative Precipitation Estimates (QPE; modification of approach in [38]) have already been tested within LIS [17]. These tests will continue as a pre-cursor to use of GPM observations to identify the optimal dataset or combination of datasets that provide the optimal data coverage.
Once near real-time data become available following the launch of the SMAP satellite, a data assimilation comparison of the three Level 2 products will be conducted. Besides product resolution, each Level 2 product is expected to have different product accuracy [39], [40], [41]. Thus, experiments must be performed to understand the trade-offs between the spatial resolution and product accuracy to determine the optimal assimilation strategy for SMAP within LIS. The most favorable simulation results from the comparison experiments will then be used to formulate the enhanced SPoRT-LIS real-time configuration. Once the GPM IMERG analysis becomes operational, an evaluation similar to that performed in [17] will also be conducted to determine the optimal strategy for ingestion of this dataset. Experiments for both the SMAP and GPM will focus on cases related to the forecast applications described in Section III.
As SPoRT development of the enhanced SPoRT-LIS product and transition to operational forecasters occurs, feedback about the lessons learned from the research and use by end-users will be communicated back to the SMAP and GPM missions to help enhance the overall application of datasets from these missions.
V. Summary
A suite of real-time, CONUS soil characteristic products that will be enhanced by inclusion of data from upcoming satellite missions SMAP and GPM has been described. These products are currently produced by the SPoRT Center through a real-time configuration of the NASA LIS and are used by operational weather forecasters to aid in analyzing and forecasting drought, extreme heat, and convective initiation. Enhancements to the SPoRT-LIS product suite will be made by assimilating Level 2 soil moisture products from SMAP to resolve local-scale features of relevance to the weather forecasting community. Also, the high temporal and spatial resolution IMERG precipitation input from GPM has the potential to improve upon current precipitation forcing datasets driving the real-time LIS integration. To expedite the transition of post-launch mission data to near real-time applications, proxy data from legacy instruments and simulated mission data will be tested to determine the optimal assimilation strategies.
Bradley T. Zavodsky has spent the past 10 years assimilating satellite-derived data into models with the goal of improving regional numerical weather forecasts. Most of this work has been done using variational assimilation schemes, such as the WRF-Var and GSI, and the WRF model.
Jonathan L. Case has 14 years of experience in technology transition, NWP models, model verification, data analysis, statistics, and mesoscale meteorology. Since 2006, he has configured, run and managed NWP and LSM experiments on parallel computing and supercomputing platforms at NASA SPoRT, incorporating unique satellite observations and products into NWP models and determining impacts on the forecast models.
Dr. Clay B. Blankenship is a research scientist for Universities Space Research Association (USRA). His research interests include satellite remote sensing, data assimilation, and numerical modeling, focused on the hydrologic cycle, the role of water vapor in the atmosphere, and interannual variability. He is currently involved with data assimilation of AIRS observations for atmospheric river events and soil moisture observations for drought and convective initiation for SPoRT and with hydrologic forecasting in East Africa for SERVIR.
Dr. William L. Crosson is a research fellow with Universities Space Research Association (USRA) at the National Space Science and Technology Center in Huntsville, Alabama. His research interests include measurement and modeling of land surface processes, especially surface-atmosphere energy exchange and the use of remotely sensed data, particularly soil moisture, to improve such models.
Kristopher D. White has 10 years of weather forecasting, analysis, and training experience in a wide range of environments, ranging from tropical maritime to continental polar. White began working for the National Weather Service (NWS) in 2006, first as a meteorologist intern at the Duluth, Minnesota, Weather Forecast Office (WFO) and from 2007 to present as a forecaster at the Huntsville, Alabama WFO.
[1] D. Entekhabi and Coauthors, “The Soil Moisture Active Passive (SMAP) mission,” Proc. IEEE, vol. 98, no. 5, pp. 704-716, 2010.
[2] Y. A. Hou, “The Global Precipitation Measurement (GPM) mission: An overview,” 2006 EUMETSAT Meteorological Satellite Conf., Helsinki, Finland, EUMETSAT, P.48. 2006.
[3] National Research Council, “Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond,” Washington, D.C.: The National Academies Press, preface., pp. xi-xvii.
[4] M. H. Freilich, “NASA Science in the Next Decade: Plans and Challenges,” 2008 American Astronautical Society National Conference and Annual Meeting, Pasadena, CA, November 2008.
[5] J. L. Case, S. V. Kumar, J. Srikishen, and G. J. Jedlovec, “Improving numerical weather predictions of summertime precipitation over the southeastern United States through a high-resolution initialization of the surface state,” Wea. Forecasting, vol. 26, no. 6, pp. 785-807, December, 2011.
[6] G. Jedlovec, “Transitioning Research Satellite Data to the Operational Weather Community: The SPoRT Paradigm,” Geoscience and Remote Sensing Newsletter, March, L. Bruzzone, editor, Institute of Electrical and Electronics Engineers, Inc., New York, pages 62-66, March 2013.
[7] S. V. Kumar and Coauthors, “Land Information System — An Interoperable Framework for High Resolution Land Surface Modeling,” Environmental Modeling & Software, vol. 21, no. 10, pp. 1402-1415, October, 2006.
[8] C. D. Peters-Lidard and Coauthors, “High-performance Earth system modeling with NASA/GSFC’s Land Information System,” Innovations Syst. Softw. Eng., vol. 3, no. 3, pp. 157-165, September, 2007.
[9] S. V. Kumar, C. D. Peters-Lidard, J. L. Eastman, and W.-K. Tao, “An integrated high-resolution hydrometeorological modeling testbed using LIS and WRF,” Environmental Modeling & Software, vol. 23, no. 2, pp. 169-181, February 2007.
[10] F. Chen and J. Dudhia, “Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity,” Mon. Wea. Rev., vol. 129, no. 4, pp. 569–585, April, 2001.
[11] M. B. Ek, K. E. Mitchell, Y. Lin, E. Rogers, P. Grunmann, V. Koren, G. Gayno, and J. D. Tarpley, “Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta Model,” J. Geophys. Res., vol. 108, no. D22, pp. 8851, November, 2003.
[12] Xia, Y., and Coauthors, “Continental-scale water and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS-2): 1. Intercomparison and application of model products,” J. Geophys. Res., vol. 117, no. D3, pp. 27, February, 2012.
[13] Y. Lin and K. E. Mitchell, “The NCEP Stage II/IV hourly precipitation analyses: Development and applications,” Preprints, 19th Conf. on Hydrology, San Diego, CA, Amer. Meteor. Soc., 1.2, January, 2005.
[14] Y. Lin, K. E. Mitchell, E. Rogers, and G. J. DiMego, “Using hourly and daily precipitation analyses to improve model water budget,” Preprints, Ninth Symp. on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface, San Diego, CA, Amer. Meteor. Soc.,3.3, January 2005.
[15] W.-S. Wu, R. J. Purser, and D. F. Parrish, “Three-dimensional variational analysis with spatially inhomogeneous covariances,” Mon. Wea. Rev., vol. 130, no. 12, pp. 2905-2916, December, 2002.
[16] Moorthi, S., H.-L. Pan, and P. Caplan, “Changes to the 2001 NCEP operational MRF/AVN global analysis/forecast system,” NWS Tech. Procedures Bull. 484, 14 pp., 2001.
[17] J. L. Case, S. V. Kumar, R. J. Kuligowski, and C. Langston, “Comparison of four precipitation forcing datasets in Land Information System simulations over the Continental U.S.,” Preprints, 27th Conf. on Hydrology, Austin, TX, Amer. Meteor. Soc., P69, January, 2013.
[18] R. A. Rozumalski, NEWR EMS User Guide. NOAA/NWS Forecast Decision Training Branch, COMET/UCAR. [Available on-line at], 2012.
[19] J. L. Case, F. J. LaFontaine, A. L. Molthan, B. T. Zavodsky, and R. A. Rozumalski, “Recent upgrades to NASA SPoRT initialization datasets for the Environmental Modeling System,” Preprints, 37th National Weather Association Annual Meeting, Madison, WI, National Weather Association, P1.40, 2012.
[20] W. C. Skamarock, J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, W. Wang, and J. G. Powers, “A Description of the Advanced Research WRF Version 2,” NCAR Tech Note. Note NCAR/TN-468+STR, pp. 88, 2005.
[21] B. Carcione, K. White, and J. L. Case, “New operational applications for the NASA Land Information System,” Abstracts, 36th Annual National Weather Association Meeting, Birmingham, AL, National Weather Association, P2.20, October 2011.
[22] K. D. White and J. L. Case, “The utility of real-time NASA Land Information System data for drought monitoring applications,” Preprints, 27th Conf. on Hydrology, Austin, TX, Amer. Meteor. Soc., P33, January 2013.
[23] K. A. Borden and S. L. Cutter, “Spatial patterns of natural hazards mortality in the United States,” International J. of Health Geographics, vol. 64, no. 7, pp. 13, December, 2008.
[24] G. Luber and M. McGeehin, “Climate change and extreme heat events.” Am. J. Prev Med. vol. 35, no. 5, pp. 429-435, November, 2008.
[25] W. T. Crow and D. Ryu, “A new data assimilation approach for improving runoff prediction using remotely-sensed soil moisture retrievals,” Hydrol. Earth Syst. Sci., vol. 13, no. 1, pp. 1–16, January, 2009.
[26] J. D. Bolten, W. T. Crow, X. Zhan, T. J. Jackson and C. A. Reynolds, 2010. Evaluating the utility of remotely sensed soil moisture retrievals for operational agricultural drought monitoring,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 3, no. 1, pp. 57-66, March, 2010.
[27] Y. Ookouchi, M. Segal, R. C. Kessler, and R. A. Pielke, “Evaluation of soil moisture effects on the generation and modification of mesoscale circulations,” Mon. Wea. Rev., vol. 112, no. 11, pp. 2281–2292, November, 1984.
[28] R. Avissar and R. A. Pielke, “A parameterization of heterogeneous land surfaces for atmospheric numerical models and its impact on regional meteorology,” Mon. Wea. Rev., vol. 117, no. 10, pp. 2113–2136, October, 1989.
[29] F. Chen and R. Avissar, “Impact of land-surface moisture variability on local shallow convective cumulus and precipitation in large-scale models,” J. Appl. Meteor., vol. 33, no. 12, pp. 1382–1401, December, 1994.
[30] E. A. Eltahir, “A soil moisture–rainfall feedback mechanism: 1. Theory and observations,” Water Resour. Res., vol. 34, no. 4, pp. 765–776, December, 1998.
[31] J. L. Case, W. L. Crosson, S. V. Kumar, W. M. Lapenta, and C. D. Peters-Lidard, “Impacts of high-resolution land surface initialization on regional sensible weather forecasts from the WRF model,” J. Hydrometeor., vol. 9, no. 6, pp. 1249-1266, December, 2008.
[32] J. M., Medlin, L. Wood, B. Zavodsky, J. L. Case, and A. Molthan, “Preliminary results of a U.S. Deep South warm season deep convective initiation modeling experiment using NASA SPoRT initialization datasets for operational National Weather Service local model runs,” Preprints, 26th Conf. on Severe Local Storms, Nashville, TN, Amer. Meteor. Soc., P121, November, 2012.
[33] S. V. Kumar, R. H. Reichle, C. D. Peters-Lidard, R. D. Koster, X. Zhan, W. T. Crow, J. B. Eylander, and P. R. Houser, “A land surface data assimilation framework using the land information system: Description and applications,” Adv. Water Res., vol. 31, no. 11, pp. 1419-1432, November, 2008.
[34] S. V. Kumar, R. H. Reichle, R. D. Koster, W. T. Crow, and C. D. Peters-Lidard, “Role of subsurface physics in the assimilation of surface soil moisture observations,” J. Hydrometeor., vol. 10, no. 6, pp. 1534-1547, December, 2009.
[35] G. J. Huffman and Coauthors, “Algorithm Theoretical Basis Document (ATBD) Version 2.0: NASA Global Precipitation Measurement (PM) Integrated Multi-satellitE Retrievals for GPM (IMERG),” 29 pp. [Available online at].
[36] K. D. McMullan, M. A. Brown, M. Martı´n-Neira, W. Rits, S. Ekhlom, J. Marti, and J. Lemanczyk, “SMOS: The payload,” IEEE Trans. Geosci. Remote Sens., vol. 46, no. 3, pp. 594–605, March, 2008.
[37] R. J. Joyce, J. E. Janowiak, P. A. Arkin, and P. Xie, “CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution,” J. Hydrometor., vol. 5, no. 3, pp. 487-503, June, 2004.
[38] R. J. Kuligowski, “A self-calibrating real-time GOES rainfall algorithm for short-term rainfall estimates,” J. Hydrometeor., vol. 3, no. 2, pp. 112-130, April, 2002.
[39] P. O’Neill and Coauthors, “Soil Moisture Active/Passive (SMAP) Algorithm Theoretical Basis Document: SMAP Level 2 & 3 Soil Moisture (Passive), version 1,” 75 pp. [Available online at], 2012.
[40] S. B. Kim and Coauthors, “Soil Moisture Active/Passive (SMAP) Algorithm Theoretical Basis Document: L2 & L3 Radar Soil Moisture (Active) Data Products, version 1,” 52 pp., [Available online at], 2012.
[41] D. Entekhabi and Coauthors, “Soil Moisture Active/Passive (SMAP) Algorithm Theoretical Basis Document: L2 & L3 Radar/Radiometer Soil Moisture (Active/Passive) Data Products, version 1,” 71 pp. [Available online at], 2012.