The Lifetime of the Covid Memorial Wall: Modelling with Collections Demography, Social Media Data and Citizen Science

TL;DR

This study demonstrates a low-cost, citizen-science–enabled collections demography workflow to model the fading and maintenance of the National Covid Memorial Wall. By combining social media-derived colour data, on-site calibration, and an agent-based model, the authors quantify repainting requirements under different strategies and paint types. They show that even with a quality paint (Paint II), sustained maintenance—approximately a few hundred hearts per week—is necessary to maintain the wall in a visually active state, underscoring the need for regular upkeep or further reductions in fading. The work highlights methodological flexibility, data limitations, and ethical considerations in using public images and volunteer data to inform heritage management. The approach provides a decision-support framework for conservation planning on large outdoor memorials and other heritage collections using publicly available data streams.

Abstract

The National Covid Memorial Wall in London, featuring over 240,000 hand-painted red hearts, faces significant conservation challenges due to the rapid fading of the paint. This study evaluates the transition to a better-quality paint and its implications for the wall's long-term preservation. The rapid fading of the initial materials required an unsustainable repainting rate, burdening volunteers. Lifetime simulations based on a collections demography framework suggest that repainting efforts must continue at a rate of some hundreds of hearts per week to maintain a stable percentage of hearts in good condition. This finding highlights the need for a sustainable management strategy that includes regular maintenance or further reduction of the fading rate. Methodologically, this study demonstrates the feasibility of using a collections demography approach, supported by citizen science and social media data, to inform heritage management decisions. An agent-based simulation is used to propagate the multiple uncertainties measured. The methodology provides a robust basis for modeling and decision-making, even in a case like this, where reliance on publicly available images and volunteer-collected data introduces variability. Future studies could improve data within a citizen science framework by inviting public submissions, using on-site calibration charts, and increasing volunteer involvement for longitudinal data collection. This research illustrates the flexibility of the collections demography framework, firstly by showing its applicability to an outdoor monument, which is very different from the published case studies, and secondly by demonstrating how it can work even with low-quality data.

Figures (10)

• Figure 1: (a) A bereaved volunteer working on repainting and re-inscribing the names of the deceased on the faded hearts, which were originally painted with red Posca markers. (b). The Friends of the Wall repaint only the hearts with legible inscriptions using premium red Valspar paint (Source: National Covid Memorial Wall UK Instagram, 21 May 2023)
• Figure 2: The main steps of a collections demography framework.
• Figure 3: The linear regression illustrates the high correlation between the L* a* b* values from the phone data and the reference measurements.
• Figure 4: The nine selected hearts that were tracked in social media images. Subsequently, the L* a* b* values (three-dimensional colour representation) and $\Delta E$ (quantitative measure of the difference between two colours) were calculated for each period corresponding to the selected hearts.
• Figure 5: Trend analysis, showing how the colour of the hearts has changed since their initial appearance in a photograph. Two hearts (3 and 4) didn't appear in enough good quality photographs to enable a regression. The time frame for Hearts 3 and 4 is September 2021 to July 2022