Widespread anthropogenic pollution of our air and waters threatens human and environmental health. Heavy metals, industrial byproducts, excess nutrients, and pharmaceuticals enter the environment through industrial effluent, agriculture, ailing infrastructure, and improper disposal. Unfortunately, the treatment of these streams is often hampered by technological limitations, implementation costs, and socioeconomic barriers. Porous heterogeneous hierarchical material (PHHM) systems offer an inexpensive and tunable way to remove a broad range of environmental contaminants. A typical PHHM is composed of a stable and porous scaffold material supporting active sites that allow for reactive mass transport. Designing materials systems requires exploring a wide range of variables without exhausting resources on unnecessary experimentation. Extensive resources have been devoted to developing techniques for high-throughput data collection, large data set interpretation, and machine learning-accelerated simulations. However, these tools are often inaccessible for the development of novel and green material systems, where costly exploration and physical laboratory experimentation remain essential. In this work, we propose a modified Design of Experiments (DOE) for use in limited-data scenarios. Additionally, we leverage experimental-computational-statistical collaborations for accelerating new materials design and improving performance in PHHM systems. Finally, we apply these tools to develop two main PHHM systems. Activated carbons/biochars produced through pyrolysis – a thermochemical conversion process conducted in an inert atmosphere – exhibit high surface area and adsorption affinity due to inherent active sites on their surface. We identified a local waste biomass product (cherry pits) and produced biochars through pyrolysis for soil amendments to prevent nutrient runoff and activated biochars as point-of-use adsorbents to remove heavy metals from drinking water. In another project, we used an experimental computational-statistical framework, to optimize an activated carbon fixed-bed reactor for industrial flue gas scrubbing and identified considerations for scaling small experimental domains to an industrial scale. We designed low-cost, sustainable polydimethylsiloxane (PDMS) foams embedded with a catalytic and photocatalytic nanocomposite. Using a modified Design of Experiments (DOE), the properties of the PDMS scaffold optimized the fluid flow and light transmittance through the system. Testing the scaffold in a plug flow reactor, we reported a degradation efficiency of 60-90% with a 60-minute retention time for the degradation of common drinking water contaminants including dyes and pharmaceuticals. This thesis highlights PHHMs and their potential as environmental remediation materials while implementing a statistically informed experimental design.