Urban greening is increasingly advocated as a strategy to counteract the loss of green spaces and associated ecosystem services due to urban densification. However, how to combine greater population density with more green spaces is still a topic of debate. Recent studies revealed cases of cities that became overall denser and greener during the last decades, but the underlying types of vegetation trends and their spatial distribution in relation to population growth have not been investigated yet.

We focus on one of the mentioned successful cases, Berlin, and apply an own-developed algorithm to examine urban vegetation dynamics using NDVI temporal series. The algorithm distinguishes between abrupt changes linked to variations in the extent of vegetation cover and gradual changes associated to vegetation growth or decline. We analyze the two dynamics between 2004 and 2017 in a 500-m circular neighborhood around the more than 332,000 residential address points in the city of Berlin, and quantify population change within the same areas.

An increase in both population density and NDVI characterized the surroundings of most of the analyzed residential address points. However, the observed NDVI increase was most frequently an effect of vegetation growth, which sometimes compensates for the loss of vegetation cover. The results question the relevance of simple NDVI-based indicators to monitor greening trends. Furthermore, they raise doubts about the greening strategies associated to densification interventions and their effectiveness in providing the ecosystem services demanded by a growing population.