The Urban Cool Valley

It should come to no surprise that as with most things, the UHI isn't the most predictable phenomenon. As the UHI refers to urban regions maintaining a warmer temperature than their neighbouring rural regions, there must be some circumstances where the situation is reversed.

If an urban region was found to be cooler than its neighbouring rural region, this is often referred to as a Negative Heat Island, primarily due to the equation for UHI being:

UHI = Urban Temperature - Rural Temperature

Also known as Urban Cool Valleys (UCV) or Urban Cool Islands (UCI), the phenomenon is very under-explored. A UCV is difficult to properly measure or come across and they don't have many negative implications compared to the UHI, which makes seeking them out both uninspiring and redundant. Nevertheless, it's could still be useful to understand them when dealing with UHI as they are technically two halves of the same phenomenon.

Formation:

There still exists no accepted consensus on the mechanisms for UCV formation. Not many studies have come across significant UCVs as most measurements are taken at times where you'd expect the UHI to be strongest (Arnfield's classification). Despite this, the UCV has been identified with some thought put into the mechanisms for formation in some UHI research expeditions. Two in particular:

Evaporative Cooling:

Proposed by Debbage and Shepherd (2015), the UCV is likely to form due to evaporative cooling within cities with poor drainage systems. Unlike rural regions, water accumulating in these cities under wet conditions would not be able to drain efficiently. As a result, thermal radiation would be weakened due to the water's high thermal heat capacity and evaporative cooling would further cool the urban regions compared to rural regions where the water would drain into the water table. 

Relative Cooling:

Proposed by Zhao et. al. (2014), the UCV is not a result of a cooling mechanism, but rather, a slower rate of warming in comparison to rural areas. As urban regions have varying surfaces that maintain high relative heat capacities, it would take more energy to raise their temperatures by a degree. During the day, the urban region is warmed more slowly than the rural regions and hence appears to be cooler when in truth, it is just lagging behind slightly.  During the evening, it would release its energy at a slower rate than the rural region hence creating a relative UHI. Due to this buffering effect slowed warming and cooling, the urban temperature range is also weaker than the rural range.

To help visualise this explanation, i've designed an illustrative chart on Matlab as seen below. 

An illustrative chart highlighting the differences in the degree at which urban and rural regions are warmed throughout the day. The urban peaks and troughs slightly lag behind and are less extreme than the rural region. UCV is seen during mid-day whilst UHI is seen during the evening. 

Neither proposition has been disproven as of yet, nor do they clash with one another. If Evaporative cooling is in fact correct, this could help with developing techniques to mitigate the UHI, whilst the Relative cooling proposition would help model the UHI development and create statistical probability charts for UHI formation using information on surface absorption potential. As of yet, this is all speculation but as mentioned earlier, the UCV could prove very useful if utilized correctly considering it, in and of itself is the most natural form of UHI mitigation.