SolutionsUseCase Wildfires atmosphere 1a

From Earth Science Information Partners (ESIP)

Return to: Use_Cases

Use Case SolutionsUseCase_Wildfires_atmosphere_1a: Local effects of Wildfires on Atmospheric Air Quality



A particular lightning/wildfire event scenario leads to a specific configuration of the data analysis and model processes needed to make effective predictions and assesements of wildfire development and movement. Some of these requirements in turn lead to subscription for detected/observed events that get published by data sources (event detection services are part of the Earth/space data injest/data systems). These data are place into into a shared event database together with a binding of the propagation specific tasks to computational processes (simulation services). Often, depending on climatic conditions (e.g. seasonal variations or wind direction reversals) it is possible that some models that are triggered by same event may have very different results and thus initiative distinctively different follow-on tasks (see

In the present scenario what is not well developed to a large extent is the initiation of computational processes/modesl using the available sensor network data, and how the model is coupled to local climatology and atmospheric sampling measurements.

Application Area Detail

Weather analysts/scientists/forecasters create event-driven scientific analysis and prediction processes - which include observed/detected events (both retrospective and current), propagation tasks that apply at the time of the event and forward in time, event related task verification and event prediction/inference tasks. Many of the tools utilized involve databases (e.g. climatologies), sensor networks, and analsyis activities and/or execution of numerical or assimilation models (with or without human involvement).

As an example: lightning in the Alaska-Yukon border area triggered numerous large forest fires in the summer of 2004, carrying smoke and other aerosols high into the atmosphere. The upwelling of particulate matter from the fires could be seen as a red plume moving across North America. For much of the summer, the particles remained high in the atmosphere and did not settle over populated areas. As the aerosols were advected into the southern US, the meteorological conditions changed: A cold front developed off the Eastern seaboard, and the dynamics of the front forced a major portion of the plume to descend to the surface and impact the air quality along the Eastern US. Episodes such as this have ramifications for human and ecosystem health and productivity on both short and longer-term time scales. (see and

Definition of Success

Through this use case, a user locates and identifies datasets (collections of related grandules) for use or processing. This process results in the user having access to a subset of the datasets in the portal that meet the requirements of the User. Individual datasets, and their constituent granules, may then be identified for further action or processing (e.g. visualization, analysis, download). As a result of analyses, sensors are tasked and/or deployed to return additional data and information.

Revision Information

Version 0.0.a

Prepared by Peter Fox


Wed Feb 07 12:01:00 MST 2007

Revision History

Modified by <Modifier Name/Affil>, <Date/time>, <Brief Description>

  • Peter Fox/HAO/ESSL/NCAR, Thu Feb 08 16:22:33 MST 2007, added data sources

Use Case Identification

Use Case Designation


Use Case Name

Local effects of Wildfires on Atmospheric Air Quality

Long description

Use Case Definition

Successful Outcomes

  • 1.

Failure Outcomes

  • 1.
  • 2.

General Diagrams

Schematic of Use case


General Requirements

Data source and model availability.

Sensor readiness and availability.

MISR and MODIS sensors record atmospheric motions of particulates and various networks of ground-based sensors record the effects of anomolous atmospheric concentrations, e.g. plumes and databases of such measurements are becoming available.

Related ground-based measures from EPA, satellite data from NASA, network data such as the National Lightning Detection Network (NLDN;, meteorological data from NOAA, and air quality model data from models such as the Community Multi-scale Air Quality (CMAQ) model must be combined to fully understand the consequences of an environmental event.

Measurements and models: MISR and MODIS sensors recorded the atmospheric motions of the particulates and. ground sensors recorded the effects of the descending plume. In particular, MISR has a unique capability to determine the height of smoke plumes using stereoscopic (multiple view angle) techniques. Recently, a group has begun using machine learning techniques to combine MISR and MODIS data and automatically identify and classify thousands of smoke plumes over North America, thereby accumulating statistics during 2004 on the geographic distribution, extent, orientation, and injection height of plumes (Averill et al., 2005). This database, being developed by two of our collaborators (Diner & Mazzoni at JPL), is directly relevant to downstream economic decision models. Ground-based measures from EPA, satellite data from NASA, meteorological data from NOAA, and air quality model data from models such as the Community Multi-scale Air Quality (CMAQ) model must be combined to fully understand the consequences of an environmental event. The high particulate measures in North Carolina during July 20-23 were undoubtedly due to the distant fire event.

Networking and data requirements: NASA satellite data (MISR, MODIS) combined with NOAA meteorological data, EPA ground station data, and CMAQ air quality models, can help us understand atmospheric particle transport and predict local ground level air quality impacts of these events. A coordinated integrated solution requires cooperation of local, county, regional, and state governments and mechanisms to enable diverse types of data to interoperate. Non-recurring data such as LIDAR measurements, which present a very narrow profile or single point measure of the air column, complement the gridded NCEP prediction model output and the EPA ground station data are available from several sources. The challenge is to combine these products to tap the knowledge (and the logical consequences of the combinations) that cannot be gleaned from any one source alone. Efforts such as the international Global Earth Observation System of Systems (GEOSS) which are linking together strategies and systems for Earth observation across multiple scientific boundaries will undoubtedly have an impact.

Use Case Elaboration (Optional for application/users)


Primary Actors

Weather analysts, scientists, forecasters.

Other Actors

Model provider, data provider. Agencies: NASA, NOAA


  • 1.
  • 2.
  • 3.


  • 1.
  • 2.
  • 3.

Normal Flow (Process Model)

  • 1. Event is initiated and ....
  • 2.
  • 3. Conditional prediction - where the conditions are queries for data

in the future and require tasking MODIS/MISR data sources or tasking the Platform (where the opportunity exists).

Pre-cursors for atmospheric disturbances are still under development and often have different timescales associated with them. On the other hand, empirical analyses for volcanic events are well established for specific volanic environments, volcano and possible eruption types and can be used to trigger the execution of either further monitoring or suggested deployment of new or additional sensors.

  • 4. Execution of the propagation tasks (that propagate observed events

via simulated models) of the process -- may lead to queing of tasks that require verification of observations or querying of collected data into the future (requiring sensor tasking).

NASA satellite data (MISR, MODIS) combined with NOAA meteorological data, EPA ground station data, and CMAQ air quality models, can help us understand atmospheric particle transport and predict local ground level air quality impacts of these events.

  • 5. Queued verification tasks may then be translated to monitoring

tasks (by an external task generator tool - part of the VSICS middleware) which are in turn scheduled for satisfaction by a task planning/scheduling tool against registered Sensor Mediators and the corresponding models of sensor operations (orbits, sensor type, sampling frequency, observation schedule, etc.). In the present scenario an example of such task is the identification and collection atmospheric sampling measurements in the projected path of the smoke plume.

  • 6. Sensor Mediators publish events that meet the subscribed and schedule

constraints. Examples of such events for our scenario(s) are propagation of dust, smoke and gases into population centers or features social infrastructure - e.g. public water sources, aircraft flight paths, limited visibility roadways, endangered species, and so on. Both types of events have associated warning levels and advance timescales for notice that are keyed to the geographic location and other details of the event.

  • 7. Published events in turn trigger execution of queued/on-hold

verification tasks. After simulation/prediction models are launched, develop the requests for data to be taken, as a validation check for the model predictions. The requests should be queued for later execution as would processes/tasks to perform the post-facto model-measurement comparison.

  • 8. Depending on the conditional science/weather analysis/prediction

plan -- the process may loop/iterate between propagate and verify -- culiminating in a revised or extended (in time) prediction. Ultimately, a well-completed analysis (1-7 above) and verification provides valuable feedback and insight to physical processes to be modeled, ability of models to capture features and predict consequences of events, etc. Thus, the entire process will be documented and recorded for later inspection.

Alternative Flows

  • 1.
  • 2.

Special Functional Requirements


Extension Points

  • SolutionsUseCase_Wildfires_atmosphere_1b: regional effects of wildfires on atmospheric air qualit=


Use Case Diagram

State Diagram

Activity Diagram

Other Diagrams

Non-Functional Requirements






Other Non-functional Requirements

Selected Technology

Overall Technical Approach


Technology A




Technology B