REWAS’ problems lie with water accounting

In my 2023 ISA paper, I point out that the water accounting (WA) schema is mainly descriptive while ISA is predictive.  When I wrote this, I was not aware that FAO and FutureWater had turned WA into a predictive Excel model. This is to be welcomed – the world needs predictive models that can compute real water savings. 

However despite this, the REWAS spreadsheet is unable to discern real water savings. This is because it is ‘stuck’ with five substantive problems originating from its parent WA framework. These five faultlines are explained below:

The system boundary of REWAS

Despite some of its supporting text discussing three scales (field, system and basin), REWAS stays with water accounting at the field scale and at the farm/system scale. It misses out what happens at the basin scale.  In my previous blog I argued that REWAS therefore could not analyse the changes in aggregate depletion which must be registered at the basin scale. 

The economics reading of WA

WA is mainly interested in accounting for dispositions of flows from withdrawals.  (REWAS doesn’t even do this particularly well).  However REWAS is essentially an economics reading of the hydrology of irrigation systems.  It is not how an irrigation engineer or agronomist sees an irrigation system.  Do not get me wrong, this is not about returning to a classical efficiency non-accounting world. But note the inputs to REWAS does not include choices that sit with irrigation practices and infrastructure. Even one of the most important non-irrigation inputs, rainfall, does not exist in WA. Similarly, its adjustment in REWAS produces no changes in paper or real savings.  See my paper for other examples. 

The water flows omitted by WA/REWAS

Because of its focus on ‘only’ the farm/system and field scales, WA omits other important ‘flows’ that mediate withdrawals and total depletion.  These are withdrawals, non-withdrawals, and reused versus unused recovered flows. Furthermore REWAS does not use other water volumes such as rainfall as a factor. It also does not use storage as an option to reduce withdrawals of scarce flows. 

The lack of dynamic algorithms in WA/REWAS

WA and REWAS do not incorporate any algorithms that shape withdrawals, final irrigated area, final aggregate depletion and final crop production.  Because of these gaps it cannot calculate field level application volumes (MCM) as compared to field level application depths (mm), and it cannot calculate pareto-checked reductions in depletion. 

Running away from irrigation efficiency

It is important that IE is not an end in itself – this was an important argument of WA. However, in a glaring omission, REWAS and WA treat irrigation efficiency as something false or dangerous.  They fail to see how irrigation efficiency is both 1) an actuator of agro-hydrological flows and 2) underwrites real water savings provided other variables are controlled.