PWS-Peak_Current-Ver_2.pdf

PERFORMANCE WORK STATEMENT

FOR

IMPROVE DETECTION OF LIGHTNING PEAK CURRENT

(13 Jun 2017)

Title: Improve Detection of Lightning Peak Current

Required Work: New lightning current sensors are being installed for 2-years on the Mobile Launch Tower (MLT) near the Vehicle Assemble Building (VAB) on NASA Kennedy Space Center (KSC). These sensors include three magnetic field change over time (B-dot) sensors and one electric field change over time (D-dot) sensor. Other related sensors may also be installed on the MLT. Use the data from those new sensors to do the following three subtasks:

Subtask-1: Develop a new method(s) to remotely detect the peak current of a cloud-to-ground return strokes with better performance than the present method used by the Mesoscale Eastern Range Lightning Information Network (MERLIN) and the National Lightning Detection Network (NLDN) and that could be implemented in MERLIN. Document performance gain.

Subtask-2: Develop a new method(s) to provide a peak current error estimate tailored to each individual return strokes. MERLIN and NLDN provide location error ellipses tailored to each stroke; provide analogous tailored error for peal current.

Subtask-3: Evaluate the performance of the rise time of cloud-to-ground lightning return stroke magnetic fields measured by the MERLIN sensors. Develop operationally useful guidelines on when the magnetic field rise time can be used as a good proxy for the return stroke peak current rise time. MERLIN measures the return stroke magnetic field rise times internally, but does not report this in its output. In addition, the rise-time detection is said to be very short-ranged, so few of the MERLIN sensors may be close enough to use strokes to the MLT to use its new lightning current sensors as ground truth. Part of this subtask should also include verifying the distance over which the MERLIN sensors can reliably detect return stroke magnetic field rise times.

Possible Approaches:

• Subtask-1: Infer the peak current from each of the ten MERLIN sensors based on the signal strength at the sensor and distance to the inferred location of the return stroke. Allowance for soil conductivity will likely be required. Given the peak currents inferred from the ten MERLIN sensors, some statistical measure of central value may be used as the estimate of the peak current, e.g. median. Or the Chi-Squared Minimization process already used for the return stroke location may provide a better estimate of peak current.

• Subtask-2: The peak currents inferred from the ten MERLIN sensors may also an estimate of the error of the peak current. If the peak currents tend to have a Gaussian distribution, then the standard deviation of the individual peak current measurements might be used. If the peak currents do not usually have a Gaussian distribution, then some range of percentiles of the measured peak currents might be used, e.g. Inter-Quartile Range for a 50% confidence interval or 2.5th to 97.5th percentile for a 95% confidence interval.

• Subtask-3: Use standard statistical performance metrics to compare to the MLT sensors ground truth to MERLIN’s magnetic field rise time, e.g. RMSE, bias, RMSE vs. return stroke intensity, bias vs. return stroke intensity, RMSE vs. distance, bias vs. distance, etc. The guidelines on when magnetic field rise time can be used as a good proxy for peak current rise time will depend on the data, but may take the form of distance limits from the return stroke to both the MERLIN sensor and the facility of interest. For operational use, simpler is usually better.

Salient Characteristics/Evaluation Criteria:

1) Experience with and knowledge of cloud-to-ground lightning return stroke peak current estimates from MERLIN or NLDN or similar locator systems.

2) Experience in comparative statistics, linear regression, non-linear regression, and correlation analysis.

Deliverables: 1) Final report summarizing the results and methodology

2) Briefing summarizing the results to 45 WS

Performance Period: Initiation Date: contract award (projected Jul 2017)

Interim Milestones: Progress reports every 6 months, or more often as needed

Completion Date: Sep 2020

Points of Contact:

Financial Contact: Name: William Roeder Organization: 45th Weather Squadron Position: Resource Advisor Phone: (321) 853-8410 E-mail: william.roeder@us.af.mil

Submitter: Name: William Roeder Organization: 45th Weather Squadron Position: Resource Advisor Phone: (321) 853-8410 E-mail: william.roeder@us.af.mil

Technical Contact: Name: William Roeder Organization: 45th Weather Squadron Position: Meteorologist Phone: (321) 853-8410 E-mail: william.roeder@us.af.mil