Charles Schramka, Massachusetts Institute of Technology; Yohan Lim, Massachusetts Institute of Technology; Carly Fridlin, Massachusetts Institute of Technology; Eric Evans, Massachusetts Institute of Technology
https://doi.org/10.64861/FWNG5813
Keywords: Space Situational Awareness, Site Selection, GIS, SSA Networking, SSA Architectures
Abstract:
As the demand for space situational awareness (SSA) data continues to grow, new optical sensors are built each year, expanding mission capabilities. A key element in growing networks of sensors is optimal placement of assets to attain maximum operational utility. This is particularly critical for optical space situational awareness, as a large variety of environmental, operational, and logistical factors impact the data quality and mission contribution of sensors. Ad-hoc placement driven by convenience or reliance on heritage locations is no longer sufficient. Increasing congestion in space traffic and fluctuating budgets necessitate that SSA sensor networks maximize operational availability and observation quality while minimizing the number of sensors.
Previous studies have been conducted using GIS tools to determine astronomical observatory placement. These methods have been used to investigate observatory sites in Antarctica, Egypt, Iran, and Turkey. However, this approach has never been undertaken for SSA or applied to networks of optical systems. Moreover, the factors influencing SSA site placement differ from those of traditional astronomical observatories. While the astronomy community may value pristine atmosphere and exquisite measurement, the SSA community most often values operational availability. When dozens of sensors are being built each year, there is insufficient time to conduct lengthy multi-year site surveys, making remote sensing and historical data approaches particularly useful.
This work presents an SSA-specific site value model, and multiple architectures built from the model are investigated using optimization techniques. The data investigated in this study include light pollution, atmospheric aerosol content, atmospheric water vapor, wind speed, fractional cloud cover, and several other environmental and logistical factors. Each dataset is processed individually to track uncertainty growth and for spatial and temporal consistency, and the combination of factors is then weighted using the analytic hierarchy process based on inputs from a panel of SSA subject-matter experts. A heatmap is then produced which indicates areas of high and low SSA site value. While scalable globally, data is processed regionally for two example problems in Peru and Alaska. Results are presented for these regions, and clear separation between areas of high and low value is noted in all cases.
An extension of the SSA site value model is to utilize historical weather data to predict the operational availability of sensor networks. In this work a genetic algorithm is applied to optimize networks of optical sensors, with the objective of maximizing operational availability subject to constraints on sensor count and observation quality. This multivariable optimization problem directly builds off the presented site value model and enables the development of robust sensor networks that balance a variety of competing factors. The results of the optimization are presented, and it is shown that a relatively small number of optical sensors can attain reasonable optical coverage in geosynchronous orbit through historical weather conditions.
While dependent on the accuracy of multiple complex datasets, and potentially biased by expert weighting, this method provides a relatively easy-to-implement alternative to ad-hoc and legacy site selection. This enables operational improvements by building sites that, following historical trends, will provide the greatest operational contribution at both the sensor observation level and the SSA network level.
Date of Conference: September 15-18, 2026
Track: Space Domain Awareness