Main Product 3 (MP3) : the DeSurvey Vulnerability Assessment Product

Authors:

Ulf Hellden (Lund University), Juan Puigdefábregas, Jaime Martinez (EEZA-CSIC)

Very often dry land management asks for very long term forecasting of the land condition rather than at the near future prediction. This us the case of vulnerability assessment of land uses systems if current conditions would persist or in other specified scenarios.

This makes sense when exploratory analyses are needed quickly and at moderate costs. This is the situation at which the DeSurvey MP3 (Land vulnerability to desertification) intends to provide answers, in doing so it will focus on specific areas where widespread cases or syndromes of desertification actually do occur.

To this purpose two complementary generic components have been developed:

A Model of the Desertification Concept (LU-CMD)
A Generic Model of Desertification (GDM)

MP3 is based on systems dynamics modelling. The first is simple enough to explore the system trajectories under several boundary conditions in areas of subsistence economy with limited data availability as to getting a sense of what desertification could involve in a range of climatic scenarios.

The second releases the subsistence economy condition and is designed to be applied as a preliminary land system's stability analysis in desertification prone areas. MP3 relies on the UN and GEF definitions of desertification. They illustrate the concept of desertification through differential equations, simulation output graphics and through causal loop diagrams demonstrating the existing feedback mechanisms.

They may be useful for Long Term land use system stability/equilibrium condition analysis and for sustainable strategic land policy and management decision support.

Finally, and as a consequence of its generic character, MP3 asks for a previous syndrome definition, that has been performed as part of its development, as a diagnostic tool to reduce the variability of its application conditions. The LU-CMD model relates population pressure and dynamics over time to the removal and availability of biomass resources.

The population stock is described as a function of growth rate, death rate and resources dependent in and out migration of people. The relative growth rate of the stock of resources is modelled as a function of climate and exploitation pressure affecting soil erosion and water availability.

Biomass recovery from serious degradation/desertification events follows the logistic growth function modified by population pressure, erosion and water availability conditions. The GDM model evolves in a procedure to build up models, which associated indicators that can be used as an early warning system of desertification.

However the notion behind proposed indicators differs from conventional ones.
These latter rely on physical measures of land degradation or environmental condition and they reflect the current state of the system; actually they are the hallmarks of a process that is in progress.

On the contrary GMD indicators are derived from the structure of equations that define the system, They are based on the intermingled economical and physical processes that characterize desertification syndromes. The objective of indicators is anticipating the impact of either current land management and climate patterns or scenarios of both, on the final stable states of the targeted systems.

These new generation of indicators should be considered as one possible utility of simulation models, the true thread of the procedure. This family of models results from a general desertification model, an eight-equation framework of a dynamic model which seeks to formally represent a broad range of desertification syndromes.

LU-CDM is a human-environment two-level (resource-man) coupled predator-prey based model. The human population is the predator and the biomass resource is the prey. The model simulates desertification over a 150 years period, 1900-2050, applying a numerical step size (delta time, DT) of
0.5 years.

It generates graphic output to illustrate the status and dynamics of all converters, system flows and stocks over time. Human population besides natural balance between birth and growth rates, includes a net migration term to account for the effects of differences of resource availability inside and outside of the target areas. Resource dynamics besides the human ‘extraction’ includes the effect of rainfall variability and soil erosion loss thru growth rate multipliers.

The model is designed to be run in subsistence economies where market and external agricultural policies have limited effects. This allows avoiding explicit economical formulations in the model.

In its present stage LU-CDM is based on the original concept of the Lotka-Volterra Predator-Prey model, further developed to simulate consumer-resource interactions.

The model is built to illustrate and test the assumption that a land use/land production system (crop land, rangeland or forest/woodland) can degrade to such an extent that it reaches a point of no return, the system 10 stability and its resilience threshold are broken, and the production system breaks down.

At that stage, the system is supposed to find a new level of equilibrium where almost no biomass is produced. The level of biomass (food, fodder, woody biomass) production becomes insufficient for human survival for
a “very long period of time”, possibly even irreversible.

A “very long period of time” is assumed to be a man age or two in a developing country. Meanwhile the affected self-subsistent population is left without a local livelihood option.
They may ultimately face famine unless they leave the area or they are assisted with imported food and other needed resources in time.

The assumed positive feedback loops that are running, or even accelerating the degradation process, are started by the initial net-removal of vegetation by humans or climate and involve as a result increased water run-off, increased soil water erosion, decreasing water infiltration to the root zone, and also increased wind erosion and soil/sand mobility for sandy soil types.

The sandy soils particle mobility makes the establishment and growth of new vegetation difficult or even impossible until the soil is stabilized through soil conservation means. The total effect of the loops is supposed to yield a positive feedback on the vegetation removal process by gradually enhancing the systems inability to support vegetation growth and cover.

This will again result in further water runoff, soil erosion…. The biomass/vegetation cover degradation rate is supposed to increase or even accelerate for every loop.

The issue of ground water exploitation and irrigation/salinization problems are excluded from the modelling approach presented here. However, in principle the biomass stock in the model can be replaced by a “water resource” stock.

The GDM relies on widely accepted partial models (i.e. the logistic growth equation of natural populations, the hill-climbing heuristic, the exponential decay of erosion against vegetation density).

Managing such a range of equations is incompatible with a very detailed process-based approach and simplifications are necessary while maintaining the requirements of both limited data availability and the possibility of quantitative analysis.

By this way MP3 is designed for application in areas with scarce data and its approach relies more on the afore-mentioned indicators that are based on the structure of a few number of simple equations, which are shown below in their generic shape MP3 is now completed in its basic scientific and technical terms.

It can be run in the areas afore mentioned under supervision of their responsible scientific teams. In the following two years of the project, it will be applied in other DeSurvey areas (i.e. Senegal, China, Maghreb and Chile).

This process will involve the participation of partners and users of these countries in such a way that technical people of the corresponding public administrations will be involved in their application. To this purpose, steps have already started including data sets collection, implementation in National Programs (Spain), design of user-oriented software, and training and demonstration.

All these activities are expected to be completed by the end of the project.