Overview
By coupling dwellings to the area of production required to support the inhabitants, we introduce a new limit to the achievable population density within a given urban area.
From the previous experiment we concluded that even as the dwellings aggregate into denser clusters by growing vertically into taller units, the population density achievable by our current production intensity is limited to 100ppl/ha. As local density increases, the overall density approaches 100ppl/ha. Using the rooftops would allow us to achieve the maximum of 100ppl/ha. Therefore in order to achieve population densities that compare to our reference cities, we will need to increase production intensity.
Production Intensity
As previously shown, our research suggests that increased productivity from a given area of land can be accomplished in a number of ways. We have broken these into four categories which we will call the productivity gradient:
- - Innovative Agricultural Techniques
- - Manipulated Production Surface (to increase surface area - includes folding, terracing, vertical/hanging plantings)
- - Enclosures which primarily use natural light (typical greenhouses) and minimally extend the growing season.
- - Highly Industrialized Enclosures which rely on artificial light and energy and can be planted year round, extending the number of harvests.
However these categories correspond to a jump in necessary capital investment and ongoing running costs associated with each type, limiting commercial viability to economic conditions. Also the higher startup costs would likely pose a barrier to entry for small farmers / entrepreneurs and lead to further consolidation of the food industry. Therefore, in order to encourage a decentralized food source system, a range of solutions is desirable.
Computational Rules
Two additional rules were introduced to allow the model to accommodate for increased population densities. Firstly, when the field is full, the algorithm allows agents to increase the production intensity of existing cells. Therefore, a settler can still build, but his/her productive needs must be met by either: an increase of yields from an existing agricultural plot, or a neighbouring built cell must introduce production, introducing ‘hybrid’ types.
Production Intensity (PI) is allowed to increase in steps of one, corresponding to number of people who’s annual fruit and vegetable needs are met per 100m2. Therefore a cell having PI = 3 means three people are provided for on that 100m2 plot.
More work needs to be done to correlate the Production Intensity values to real world implementations, however our research has indicated that PI=3 could be achieved with readily available greenhouse technology, PI=5 is achievable with highly specialized greenhouse technology, and PI=10 could achieved by a two level high-tech energy intensive greenhouse.
The second rule relates to ‘Aggressiveness of Development’ (AD) and is implemented to allow the user to adjust the likelihood that each built cell be adjacent to at least one productive cell. Therefore when Density and Friendliness are high, setting AD = 0 ensures that all inhabitants live adjacent to at least one agricultural plot.
(F) Friendliness Factor: F= 0.3 , 0.6, 0.9
(D) Population Density: D= 140, 220, 280
(Ag) Aggressiveness of Development: 0 to 1 (variable)
(PC) Productivity Ceiling = 10
(HC) Height Max (Allowable # of Floors) = 6
Qualities Measured
Method
The revised Settlement Simulation was run for a variety of parameter settings to compare the combined effects of Friendliness, Density and Aggressiveness on Production Intensity and urban qualities such as Open Space Ratio.
Thirty tissues were generated using three Density targets each run with ten F values and AD=0. Nine were selected and with those F and D values run again with AD=1.
Population Densities used were based on sample tissues studied. 280 ppl/ha is the density of Gracia neighbourhood in Barcelona, 220 ppl/ha relates to Brooklyn Heights in Brooklyn, NY and 140 ppl/ha corresponds to Stoke Newington in London.
Generated tissues were studied for the resulting Productivity Gradient. Frequency distributions were plotted for the resulting productivity values giving us an idea of what the population density threshold might be for a system which is more conducive to small farmers and entrepreneurs.
The effects of Density and Friendliness work as opposing forces; Density fills the field by increasing the amount of material that can be ‘poured’ in. Without the Friendliness Factor (F=0) the simulation results in a Gaussian distribution of material or a maximally disordered state.
Friendliness then acts much as a centripetal force, pulling the dwellings into tightly packed ‘piles’ at certain attractor points, the locations of which are the determined by the first few dwellings that are placed. (The earliest placed dwellings create the largest aggregations.)
In this way, Friendliness has the effect of introducing order to the otherwise random state space.
Production cells are pushed out of the dwelling clusters, but the algorithm dictates that they stay close. Therefore large clusters of dwellings result in large aggregations of high intensity production gathered around the perimeter of their volumes.
While ‘D’ is high and ‘F’ is low, we see anomalies in the distribution of Production Intensities.
In the PI distributions for sample tissues: F=0.45|D=220, F=0.45|D=280, and F=0.85|D=280 we see a drop in the frequency of PI=3 units, then disproportionately more high-intensity cells compared with the other samples. As the friendliness is increased for each of these density values, the distributions return to a pattern of diminishing frequencies, similar to the other tissues.
We believe this is likely attributed to low Open Space Ratios, which result from high D with low F. In these tissues, dwelling clusters overwhelm open plots and settlers begin aggregating high-intensity productive units around themselves, placing their own production atop their neighbours. Therefore, when an economical distribution of production intensities is considered desirable, the following thresholds should be considered.
Aggressiveness of Development (AD) has little effect on the distribution of Production Intensity types.
OSR is similar in both sets of samples, although the size of ‘continuous’ open plots is reduced slightly. The compromise here could be weighed against the benefits resulting from the dwelling clusters having more direct access to small agricultural plots. This could be considered advantageous in a scenario which calls for a greater percentage of the population ‘growing their own’ while still requiring high densities. This can be studied in greater detail once cluster sizes can be identified in the samples (see following sections).
In addition, ‘livability’ of the dwelling clusters could be considered improved when AD=0, by ensuring better light, ventilation and views by the penetration of agricultural cells within the large dwelling clusters.
(D) Population Density: D= 140, 220, 280
(Ag) Aggressiveness of Development: 0 to 1 (variable)
(PC) Productivity Ceiling = 10
(HC) Height Max (Allowable # of Floors) = 6
Qualities Measured
(PG) Productivity Gradient: Frequency Distribution of Production Intensities from 0 to 9.
(OSR) Open Space Ratio: OSR = (1 - C) / FAR
- where:
- Coverage (C) = Footprint / Sample Area
- and
- Floor Area Ratio (FAR) = Gross Floor
- Area / Sample Area.
Method
The revised Settlement Simulation was run for a variety of parameter settings to compare the combined effects of Friendliness, Density and Aggressiveness on Production Intensity and urban qualities such as Open Space Ratio.
Thirty tissues were generated using three Density targets each run with ten F values and AD=0. Nine were selected and with those F and D values run again with AD=1.
Population Densities used were based on sample tissues studied. 280 ppl/ha is the density of Gracia neighbourhood in Barcelona, 220 ppl/ha relates to Brooklyn Heights in Brooklyn, NY and 140 ppl/ha corresponds to Stoke Newington in London.
Generated tissues were studied for the resulting Productivity Gradient. Frequency distributions were plotted for the resulting productivity values giving us an idea of what the population density threshold might be for a system which is more conducive to small farmers and entrepreneurs.
Observations & Discussion
The introduction of Production Intensity rules allowed us to achieve the desired Density targets while meeting productivity requirements for that population.The effects of Density and Friendliness work as opposing forces; Density fills the field by increasing the amount of material that can be ‘poured’ in. Without the Friendliness Factor (F=0) the simulation results in a Gaussian distribution of material or a maximally disordered state.
Friendliness then acts much as a centripetal force, pulling the dwellings into tightly packed ‘piles’ at certain attractor points, the locations of which are the determined by the first few dwellings that are placed. (The earliest placed dwellings create the largest aggregations.)
In this way, Friendliness has the effect of introducing order to the otherwise random state space.
Production cells are pushed out of the dwelling clusters, but the algorithm dictates that they stay close. Therefore large clusters of dwellings result in large aggregations of high intensity production gathered around the perimeter of their volumes.
While ‘D’ is high and ‘F’ is low, we see anomalies in the distribution of Production Intensities.
In the PI distributions for sample tissues: F=0.45|D=220, F=0.45|D=280, and F=0.85|D=280 we see a drop in the frequency of PI=3 units, then disproportionately more high-intensity cells compared with the other samples. As the friendliness is increased for each of these density values, the distributions return to a pattern of diminishing frequencies, similar to the other tissues.
We believe this is likely attributed to low Open Space Ratios, which result from high D with low F. In these tissues, dwelling clusters overwhelm open plots and settlers begin aggregating high-intensity productive units around themselves, placing their own production atop their neighbours. Therefore, when an economical distribution of production intensities is considered desirable, the following thresholds should be considered.
- - For D=140 use F > 0.45
- - For D=220 use F > 0.85
- - For D=280 use F > 0.95
Aggressiveness of Development (AD) has little effect on the distribution of Production Intensity types.
OSR is similar in both sets of samples, although the size of ‘continuous’ open plots is reduced slightly. The compromise here could be weighed against the benefits resulting from the dwelling clusters having more direct access to small agricultural plots. This could be considered advantageous in a scenario which calls for a greater percentage of the population ‘growing their own’ while still requiring high densities. This can be studied in greater detail once cluster sizes can be identified in the samples (see following sections).
In addition, ‘livability’ of the dwelling clusters could be considered improved when AD=0, by ensuring better light, ventilation and views by the penetration of agricultural cells within the large dwelling clusters.
