National Weather Service Office Research Articles

NOAA's National Water Model: Advancing operational hydrology through continental‐scale modeling

AbstractThe National Weather Service (NWS) Office of Water Prediction (OWP), in conjunction with the National Center for Atmospheric Research and the NWS National Centers for Environmental Prediction (NCEP) implemented version 2.1 of the National Water Model (NWM) into operations in April of 2021. As with the initial version implemented in 2016, NWM v2.1 is an hourly cycling analysis and forecast system that provides streamflow guidance for millions of river reaches and other hydrologic information on high‐resolution grids. The NWM provides complementary hydrologic guidance at current NWS river forecast locations and significantly expands guidance coverage and water budget information in underserved locations. It produces a full range of hydrologic fields, which can be leveraged by a broad cross section of stakeholders ranging from the emergency responder and water resource communities, to transportation, energy, recreation and agriculture interests, to other water‐oriented applications in the government, academic and private sectors. Version 2.1 of the NWM represents the fifth major version upgrade and more than doubles simulation skill with respect to hourly streamflow correlation, Nash Sutcliffe Efficiency, and bias reduction, over its original inception in 2016. This paper will discuss the driving factors underpinning the creation of the NWM, provide a brief overview of the model configuration and performance, and discuss future efforts to improve NWM components and services.

Combining Probabilistic Hazard Information Forecast Graphics with Wireless Emergency Alert Messages: An Exploratory, Qualitative Study

Abstract Since 2019, National Weather Service (NWS) offices have been able to issue 360-character Wireless Emergency Alert (“WEA360”) messages for tornadoes. NWS is now considering changing from a “deterministic” to a “probabilistic” warning paradigm. That change could possibly influence how WEA360 messages for tornado are issued in the future. Recent experimental studies have found that probabilistic hazard information (PHI) forecast graphics improve consumers’ risk perception for tornadoes, but findings from these studies concerning whether PHI forecast graphics improve people’s protective action decision-making are mixed. The present study therefore investigated how mock PHI-enhanced WEA360 messages might influence people’s risk perception and protective action decision-making. Analysis of qualitative data gathered from a combination of questionnaire and focus group interview methods conducted in collaboration with 31 community members in Denver, Colorado, indicated that inclusion of PHI forecast graphics within WEA360 messages elicited high levels of understanding and message believability but did not consistently lead to appropriate precautionary intent. Because warning response is a complex social phenomenon, PHI may not significantly improve protective action decision-making if PHI forecast graphics are eventually presented to consumers via the Wireless Emergency Alerts system. Factors that PHI stakeholders should consider before the adoption of PHI-enhanced WEA360 messages for consumers are discussed. Significance Statement This study examines how consumers respond to and talk about mock WEA360 messages for tornadoes that contain embedded PHI forecast graphics. As NWS considers moving to a probabilistic warning paradigm, stakeholders will need to determine how PHI forecast graphics might be communicated directly to consumers, if at all. Our findings suggest that combining WEA360 messages with PHI forecast graphics creates challenges and complexities related to consumers’ assessment of personal risk and protective action decision-making. Overall, the study suggests that any future PHI-enhanced WEA360 messages provided directly to consumers, if at all, must avoid discrepancies (even subtle) between the level of risk represented by the PHI forecast graphic and the protective action guidance included in the text of the messages.

Prospects for Machine Learning Activity within the United States National Weather Service

Abstract The National Weather Service (NWS) Office of Science and Technology Integration commissioned a report to assess the status of artificial intelligence (AI) and machine learning (ML) activity within the agency with a view toward identifying existing obstacles and recommending future directions. The purpose of this essay is to communicate the steps that the NWS plans to take to realize the potential benefits of AI in operations. AI activities are growing rapidly within atmospheric sciences, and the NWS is part of this growth. However, the activity is fragmented and lacks the needed infrastructure for improved coordination of effort. Current obstacles to progress include insufficient workforce training in AI/ML, a lack of curated datasets and software that can be used for development and evaluation of these approaches, the absence of a centralized clearing house available to NWS personnel for technical expertise and consultation, limited operational compute resources, and a lack of a clear end-to-end project pathway that encompasses exploration, development, testbed/proving ground, and operational implementation. These limitations are addressable. Training materials specific to NWS interests can be developed through collaboration with existing NOAA centers. Establishing a reference library staffed with AI/ML consultants tasked with collaborating with operational units would reduce siloed efforts and enhance productivity. Establishing funding vehicles for theme-based projects with a sustainable pathway through operational implementation would help bridge the research-to-operations “valley of death.” Given the growth of AI/ML across the U.S. Weather Enterprise and the already substantial involvement of academic and private sector entities, these developments within the NWS will be of interest to the atmospheric science field.

Sharpiegate

In this article, we use the case of Hurricane Dorian, which hit the southeastern United States in 2019, to expand on the scholarship about professional vision by taking up how weather is materialized in the negotiation of seeing between members of the scientific community and the state. Attending to the controversy known popularly as Sharpiegate, in which President Donald Trump, the National Weather Service (NWS) office in Birmingham (Alabama) and the National Oceanic and Atmospheric Administration (NOAA) offered conflicting accounts of Hurricane Dorian’s predicted path, we use discourse analysis to examine four tweets, a NOAA official statement and the Commerce Department’s investigative report to show how institutional actors negotiated their entitlement to see, accounts of what was seen and accountability to their claims as seers. We argue that Sharpiegate significantly disrupted the relationship between science and the state, which depends upon scientists’ professional re-semiotizing of hurricanes in images which the state can use to legitimize its crisis response. Our findings show how NOAA and the Commerce Department utilize strategic ambiguity, distributed agency and accountability and the careful maintenance of boundaries in their claims of what was seen and what counts as legitimate seeing in weather forecasting.

Backyard Hydroclimatology: Citizen Scientists Contribute to Drought Detection and Monitoring

Abstract The Community Collaborative Rain, Hail and Snow (CoCoRaHS) network is a well-regarded, trusted source of precipitation data. The network’s volunteers also provide weather and climate observations through daily comments, significant weather reports, and condition monitoring reports. Designed to meet a need for local information about drought events and their impacts, “condition monitoring” was initiated as a pilot project in North Carolina and South Carolina in 2013 and launched nationally in October 2016. Volunteers regularly report on how precipitation, or a lack thereof, affects their local environment and community by ranking current conditions on a seven-point scale ranging from severely dry to severely wet and sharing observations through written narratives. This study assesses the usefulness of these reports for drought monitoring and decision-making, drawing from the >7,100 reports submitted in the Carolinas between October 2016 and June 2020. This period encompasses the Carolinas’ climate patterns and extreme events such as droughts, wildfires, and hurricanes (“drought busters”). Three aspects of usefulness were evaluated in the reports: the extent to which volunteers’ assessments of dry-to-wet conditions correspond to objective drought indicators (EDDI, SPI, SPEI) typically employed for monitoring drought; how volunteers’ qualitative observations depict changing conditions, focusing on two flash droughts in 2019; and actual use of the reports by National Weather Service offices, State Climate Offices, U.S. Drought Monitor authors, and drought response committees. Although ­report content can vary widely, findings show that volunteers’ assessments reflect meteorological conditions and provide on-the-ground details that are being incorporated into existing drought monitoring processes.

Identifying community values related to heat: recommendations for forecast and health risk communication

Abstract. Effective communication of heat risk to public audiences is critical for promoting behavioral changes that reduce susceptibility to heat-related illness. The U.S. National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) provides heat-related information to the public using social media platforms such as Facebook. We applied a novel rhetorical framework to evaluate 5 years (2015–2019) of public responses to heat-related Facebook posts from the NWS office in Phoenix (Arizona) to identify “commonplaces” or community norms, beliefs, and values that may present challenges to the effectiveness of heat risk communication. Phoenix is in one of the hottest regions in North America and is the 10th-largest metropolitan area in the U.S. We found the following two key commonplaces: (1) the normalization of heat and (2) heat as a marker of community identity. These commonplaces imply that local audiences may be resistant to behavioral change, but they can also be harnessed in an effort to promote protective action. We also found that public responses to NWS posts declined over the heat season, further suggesting the normalization of heat and highlighting the need to maintain engagement. This work provides a readily generalizable framework for other messengers of high-impact weather events to improve the effectiveness of their communication with receiver audiences.

A Coastal Flood Event Database for the Southeastern Georgia and Southeastern South Carolina Coast and the Operational Implementation of a Tide Forecast Tool

Coastal flooding occurs when saltwater inundates normally dry land and the resulting impacts can range from minor flooding of low-lying areas along the coast, to significant damage to property and structures. Previous research consistently suggests that if sea-level rise continues to increase along the East Coast of the United States, coastal flooding will occur more frequently. In order to document the history of coastal flooding along the southeastern Georgia and southeastern South Carolina coast, a coastal flood event database was created for National Ocean Service tide gauges located in Charleston Harbor, South Carolina and Fort Pulaski, Georgia. Trends from the data show that coastal flooding is occurring more frequently with time at both tide gauges, particularly over the last five to ten years. Because of the increased frequency and worsening impacts of tidal flooding, a tide forecast tool is implemented operationally in an effort to improve deterministic tide forecasts. This study extends the dataset used in the Charleston Harbor forecast tool, expands the tool to Fort Pulaski, and compares the synoptic category forecast equations to an all-inclusive equation that does not differentiate by synoptic category. Results show that there is virtually no difference in the forecast accuracy between the all-inclusive forecast equation and the specific forecast equations based on synoptic category. Furthermore, the all-inclusive forecast equation can be implemented operationally, will help improve deterministic tide forecasts, and will likely aid in the decision-making process for Coastal Flood Watches, Warnings, and Advisories issued by the National Weather Service office in Charleston, South Carolina.

Evaluating Multi-Radar, Multi-Sensor Products for Surface Hailfall Diagnosis

The operational deployment of the multi-radar, multi-sensor (MRMS) system has made available new products to use for hail detection. MRMS products are provided on a spatial grid and can give information on hail size and the spatial extent and distribution of the hail fall. This information is important to a wide audience, including warning forecasters needing to focus on areas for warning verification and insurance users needing to verify a claim. Products are typically verified and evaluated using hail reports from Storm Data, which are reports collected by local National Weather Service Offices. The National Severe Storms Laboratory conducted a project to collect reports of hail, including reports of “no hail” near storms, at high spatial resolution. This project, the Severe Hazards Analysis and Verification Experiment (SHAVE), collected tens of thousands of hail reports over ten years of operations. Three-hundred eighty-nine SHAVE operations, which yielded 21 184 SHAVE reports and 2814 Storm Data reports, are investigated. Nine MRMS products were evaluated with the reflectivity at lowest altitude demonstrating the best discrimination for where hail of any size fell and the maximum expected size of hail product provided the best discrimination for severe-sized hail. SHAVE- and Storm Data-based evaluations showed marked differences in product skill scores. Discussions on the differences between the hail report databases and explorations of vertical profiles of reflectivity are included.

Evaluation of NOAA National Water Model Parameter Calibration in Semi-Arid Environments Prone to Channel Infiltration

AbstractThe NOAA National Water Model (NWM), maintained and executed by the NOAA National Weather Service (NWS) Office of Water Prediction, provides operational hydrological guidance throughout the Contiguous United States. Based on the WRF-Hydro model architecture developed by the National Center for Atmospheric Research (NCAR), the NWM was recently modified for semi-arid domains, by permitting it to explicitly resolve infiltration from ephemeral channels into the underlying channel bed as an added model sink term. To analyze the added value of channel infiltration in semi-arid environments, we calibrated NWM v2.1 (with the channel infiltration function) to 56 independent basins in the western CONUS, following identical calibration methods as the pre-operational NWM v2.1 (not including channel infiltration). Calibration of the model consists of two parts, including 1) calibration of channel infiltration only with other parameters set to the calibrated parameters used for pre-operational NWM v2.1 and 2) calibration of all parameters including channel infiltration with settings otherwise equivalent to the calibration of NWM v2.1. The calibrated channel-infiltration enhanced NWM improves predictive skill compared to the control NWM in 85% of evaluated basins, for the calibration period. The current NWM settings for physical processes and the biases of the calibration scheme limit model performance in semi-arid environments. To explore whether channel infiltration paired with an alternative calibration scheme could address these limitations, NWM v2.1 was calibrated with a new objective function in selected basins. We found that this updated objective function could ameliorate model biases in some semi-arid environments.

The Precipitation Imaging Package: Phase Partitioning Capabilities

Surface precipitation phase is a fundamental meteorological property with immense importance. Accurate classification of phase from satellite remotely sensed observations is difficult. This study demonstrates the ability of the Precipitation Imaging Package (PIP), a ground-based, in situ precipitation imager, to distinguish precipitation phase. The PIP precipitation phase identification capabilities are compared to observer records from the National Weather Service (NWS) office in Marquette, Michigan, as well as co-located observations from profiling and scanning radars, disdrometer data, and surface meteorological measurements. Examined are 13 events with at least one precipitation phase transition. The PIP-determined onsets and endings of the respective precipitation phase periods agree to within 15 min of NWS observer records for the vast majority of the events. Additionally, the PIP and NWS liquid water equivalent accumulations for 12 of the 13 events were within 10%. Co-located observations from scanning and profiling radars, as well as reanalysis-derived synoptic and thermodynamic conditions, support the accuracy of the precipitation phases identified by the PIP. PIP observations for the phase transition events are compared to output from a parameterization based on wet bulb and near-surface lapse rates to produce a probability of solid precipitation. The PIP phase identification and the parameterization output are consistent. This work highlights the ability of the PIP to properly characterize hydrometeor phase and provide dependable precipitation accumulations under complicated mixed-phase and rain and snow (or vice versa) transition events.

A Composite Analysis of Snowfall Modes from Four Winter Seasons in Marquette, Michigan

AbstractPresented are four winter seasons of data from an enhanced precipitation instrument suite based at the National Weather Service (NWS) Office in Marquette (MQT), Michigan (250–500 cm of annual snow accumulation). In 2014 the site was augmented with a Micro Rain Radar (MRR) and a Precipitation Imaging Package (PIP). MRR observations are utilized to partition large-scale synoptically driven (deep) and surface-forced (shallow) snow events. Coincident PIP and NWS MQT meteorological surface observations illustrate different characteristics with respect to snow event category. Shallow snow events are often extremely shallow, with MRR-indicated precipitation heights of less than 1500 m above ground level. Large vertical reflectivity gradients indicate efficient particle growth, and increased boundary layer turbulence inferred from observations of spectral width implies increased aggregation in shallow snow events. Shallow snow events occur 2 times as often as deep events; however, both categories contribute approximately equally to estimated annual accumulation. PIP measurements reveal distinct regime-dependent snow microphysical differences, with shallow snow events having broader particle size distributions and comparatively fewer small particles and deep snow events having narrower particle size distributions and comparatively more small particles. In addition, coincident surface meteorological measurements indicate that most shallow snow events are associated with surface winds originating from the northwest (over Lake Superior), cold temperatures, and relatively high surface pressures, which are characteristics that are consistent with cold-air outbreaks. Deep snow events have meteorologically distinct conditions that are accordant with midlatitude cyclones and frontal structures, with mostly southwest surface winds, warmer temperatures approaching freezing, and lower surface pressures.

More hots: Quantifying upward trends in the number of extremely hot days and nights in Tallahassee, Florida, USA: 1892–2018

AbstractHot day and night occurrences in Tallahassee, Florida, USA are analysed and modelled. A hot day is defined as one during which the high temperature exceeds 100°F (37.8°C). A hot night is defined as one during which the low temperature fails to drop below 77°F (25°C). The U.S. National Weather Service Office (WSO) Tallahassee official record shows an upward trend in the number of hot days at a rate of 2.1% (±.96% margin of error [moe]) per year and a more pronounced upward trend in the number of hot nights at a rate of 4.5% (±.71% moe) per year. Increasingly frequent hot days and nights result from more and longer hot events (consecutive hot days/nights). Upward trends estimated from a 127‐year time series of annual hot day/night counts, with the years prior to 1940 adjusted for location, are consistent with upward trends estimated over the shorter, more recent, period. With projected continued warming we expect more hot days and nights making uncomfortable and unhealthy conditions more common in the city.

Public understanding of local tornado characteristics and perceived protection from land-surface features in Tennessee, USA.

Misunderstandings about the influence of land-surface features on tornado frequency and other tornado-related misconceptions may affect how people prepare for and behave during hazardous weather events. This research uses a phone survey (n = 1804) to assess how participants in three regions of Tennessee perceive their local tornado characteristics (i.e., direction of travel, seasonality, and diurnal timing) and their belief in protection from land-surface features (i.e., hills, water bodies, and buildings). Region of residence influences most beliefs in local tornado characteristics, and demographic characteristics, specifically age and gender, also have some influence. Residents in hilly East Tennessee are more likely to believe they are protected by hills and underestimate the proportion of nocturnal tornadoes, while residents in West Tennessee are more likely to believe they are protected by water bodies, perhaps because of proximity to the Mississippi River. Outside of the typical severe-weather season, participants were uncertain of when tornadoes were likely to occur; specifically, they did not recognize their local wintertime tornado activity. Because public perceptions are related to local features, local organizations and personnel, for example National Weather Service offices and broadcast meteorologists, may be most helpful in dispelling these misconceptions.

Build community before the storm: The National Weather Service's social media engagement

AbstractThe current study examines crisis communication on social media by observing how twelve National Weather Service (NWS) offices use Twitter to facilitate engagement with stakeholders during threat and nonthreat periods. Using content analytic methods, we examine message features related to content and structure during a 3‐month period in spring 2016. We conduct chi‐square analyses to determine how the prevalence of these features varies by time. Results indicate that NWS offices use Twitter for ongoing engagement communication through both the content and structure of their messages, where community building and action‐orientated messages are primarily used during nonthreat periods. When fair weather changes and storms approach, the offices shift to the communication of risk. Our findings underscore the need to examine organizational communication practices in a manner that recognizes the distinctions between long‐term engagement goals and short‐term goals related to the mission of the organization.

Enhancing the Structure of the WRF-Hydro Hydrologic Model for Semiarid Environments

Abstract In August 2016, the National Weather Service Office of Water Prediction (NWS/OWP) of the National Oceanic and Atmospheric Administration (NOAA) implemented the operational National Water Model (NWM) to simulate and forecast streamflow, soil moisture, and other model states throughout the contiguous United States. Based on the architecture of the WRF-Hydro hydrologic model, the NWM does not currently resolve channel infiltration, an important component of the water balance of the semiarid western United States. Here, we demonstrate the benefit of implementing a conceptual channel infiltration function (from the KINEROS2 semidistributed hydrologic model) into the WRF-Hydro model architecture, configured as NWM v1.1. After calibration, the updated WRF-Hydro model exhibits reduced streamflow errors for the Walnut Gulch Experimental Watershed (WGEW) and the Babocomari River in southeast Arizona. Model calibration was performed using NLDAS-2 atmospheric forcing, available from the NOAA National Centers for Environmental Prediction (NCEP), paired with precipitation forcing from NLDAS-2, NCEP Stage IV, or local gauge precipitation. Including channel infiltration within WRF-Hydro results in a physically realistic hydrologic response in the WGEW, when the model is forced with high-resolution, gauge-based precipitation in lieu of a national product. The value of accounting for channel loss is also demonstrated in the Babocomari basin, where the drainage area is greater and the cumulative effect of channel infiltration is more important. Accounting for channel infiltration loss thus improves the streamflow behavior simulated by the calibrated model and reduces evapotranspiration bias when gauge precipitation is used as forcing. However, calibration also results in increased high soil moisture bias, which is likely due to underlying limitations of the NWM structure and calibration methodology.

Perceptions about Social Science among NWS Warning Coordination Meteorologists

AbstractNOAA has recently placed greater emphasis on implementing social science findings into its products, but perceptions of social science research among National Weather Service offices have n...

Restructuring the National Weather Service

Restructuring the National Weather Service

Assessing the Efficacy of High-Resolution Satellite-Based PERSIANN-CDR Precipitation Product in Simulating Streamflow

AbstractThis study aims to investigate the performance of Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks–Climate Data Record (PERSIANN-CDR) in a rainfall–runoff modeling application over the past three decades. PERSIANN-CDR provides precipitation data at daily and 0.25° temporal and spatial resolutions from 1983 to present for the 60°S–60°N latitude band and 0°–360° longitude. The study is conducted in two phases over three test basins from the Distributed Hydrologic Model Intercomparison Project, phase 2 (DMIP2). In phase 1, a more recent period of time (2003–10) when other high-resolution satellite-based precipitation products are available is chosen. Precipitation evaluation analysis, conducted against stage IV gauge-adjusted radar data, shows that PERSIANN-CDR and TRMM Multisatellite Precipitation Analysis (TMPA) have close performances with a higher correlation coefficient for TMPA (~0.8 vs 0.75 for PERSIANN-CDR) and almost the same root-mean-square deviation (~6) for both products. TMPA and PERSIANN-CDR outperform PERSIANN, mainly because, unlike PERSIANN, TMPA and PERSIANN-CDR are gauge-adjusted precipitation products. The National Weather Service Office of Hydrologic Development Hydrology Laboratory Research Distributed Hydrologic Model (HL-RDHM) is then forced with PERSIANN, PERSIANN-CDR, TMPA, and stage IV data. Quantitative analysis using five different statistical and model efficiency measures against USGS streamflow observation show that in general in all three DMIP2 basins, the simulated hydrographs forced with PERSIANN-CDR and TMPA have close agreement. Given the promising results in the first phase, the simulation process is extended back to 1983 where only PERSIANN-CDR rainfall estimates are available. The results show that PERSIANN-CDR-derived streamflow simulations are comparable to USGS observations with correlation coefficients of ~0.67–0.73, relatively low biases (~5%–12%), and high index of agreement criterion (~0.68–0.83) between PERSIANN-CDR-simulated daily streamflow and USGS daily observations. The results prove the capability of PERSIANN-CDR in hydrological rainfall–runoff modeling application, especially for long-term streamflow simulations over the past three decades.

The extension of lightning flashes from thunderstorms near Cape Canaveral, Florida

AbstractThe horizontal distance that intracloud (IC) and cloud to ground (CG) lightning flashes extend from thunderstorms is presented for warm season nonsevere storms near Cape Canaveral, FL. The Warning Decision Support System–Integrated Information (WDSS‐II) software allowed us to ingest, display, and make distance calculations using radar reflectivity data from nearby National Weather Service Offices and lightning information from the National Lightning Detection Network and the local Lightning Detection and Ranging network. The flash distances were determined manually using a measuring tool within WDSS‐II. Six categories of flashes were investigated: (a) CG flashes measured from the edge of the convective core (30 dBZ) to the ground strike location that is beneath the attached anvil, (b) CG “bolt from the blue” flashes measured from cloud edge (0 dBZ) to the ground strike point that contains only 0 dBZ overhead, (c) CG bolt from the blue flashes as in Category B but measured from the edge of the precipitation (18 dBZ), (d) IC flashes that exit a portion of the cloud that is not an anvil, measured from cloud edge (0 dBZ), (e) IC flashes within anvils measured from the edge of convective precipitation (30 dBZ) to the termination point within the anvil, and (f) IC flashes exiting the edge (0 dBZ) of anvils. Results for each category are presented in the forms of frequency distributions of distances and their cumulative frequency distributions.

CREST-Snow Field Experiment: analysis of snowpack properties using multi-frequency microwave remote sensing data

Abstract. The CREST-Snow Analysis and Field Experiment (CREST-SAFE) was carried out during January–March 2011 at the research site of the National Weather Service office, Caribou, ME, USA. In this experiment dual-polarized microwave (37 and 89 GHz) observations were accompanied by detailed synchronous observations of meteorology and snowpack physical properties. The objective of this long-term field experiment was to improve understanding of the effect of changing snow characteristics (grain size, density, temperature) under various meteorological conditions on the microwave emission of snow and hence to improve retrievals of snow cover properties from satellite observations. In this paper we present an overview of the field experiment and comparative preliminary analysis of the continuous microwave and snowpack observations and simulations. The observations revealed a large difference between the brightness temperature of fresh and aged snowpack even when the snow depth was the same. This is indicative of a substantial impact of evolution of snowpack properties such as snow grain size, density and wetness on microwave observations. In the early spring we frequently observed a large diurnal variation in the 37 and 89 GHz brightness temperature with small depolarization corresponding to daytime snowmelt and nighttime refreeze events. SNTHERM (SNow THERmal Model) and the HUT (Helsinki University of Technology) snow emission model were used to simulate snowpack properties and microwave brightness temperatures, respectively. Simulated snow depth and snowpack temperature using SNTHERM were compared to in situ observations. Similarly, simulated microwave brightness temperatures using the HUT model were compared with the observed brightness temperatures under different snow conditions to identify different states of the snowpack that developed during the winter season.