# Transport Module (40_transport)¶

**Table of contents**- Transport Module (40_transport)

## Description¶

The transport module simulates the intraregional transportation of agricultural products between producer site and the next city centre (market). It covers the transport of inputs such as fertilizers to the production site as well as the transport of products to the market. Calculations are based on the cellular production patterns which are provided by the core to the module. Returned are the corresponding cellular transport costs which are used by the core as well as by 90_presolve.

## Interfaces¶

### Input¶

Name Description Unit A B $vm\_prod(j,k)$ production in each cell mio. ton DM x

_{The last columns of the table indicate the usage in the different realizations (numbered with capital letters)}

### Output¶

Name Description Unit $vm\_cost\_transp(j,k)$ transportation costs mio. US$

### Interface plot¶

Figure 0: Information exchange among modules

## Realizations¶

### (A) on *(default)*¶

In this realization transportation costs are calculated based on the assumption that transport costs are proportional to the mass which has to be transported and the time which is required for the transport.

Equation 1:

\begin{align}

vm\_cost\_transp(j,k) \ge vm\_prod(j,k) \cdot f40\_distance(j) \cdot f40\_transport\_costs(k)

\end{align}

Correspondingly, the transport costs $vm\_cost\_transp(j,k)$ for each cell $j$ are calculated as the product of the production $vm\_prod(j,k)$ of the commodity $k$ and the transport distance $f40\_distance(j)$ to the next city centre, measured in time and the estimated transportation costs per time and mass $f40\_transport\_costs(k)$.

As cellular distance information $f40\_distance(j)$, the European Commission Joint Research Centre (EC JRC)’s 30 arc-second resolution map on travel time for any location on the earth surface to the nearest large city is used (Figure 1 - Nelson^{1} 2008). The data set is based on multiple indicators (biophysical, administrative and transport mode) which determine the friction surface that in turn determines the speed needed to transport goods across grid cells. The cumulated time value needed to reach an urban center of minimal 50000 inhabitants stands as static proxy for accessibility of a grid cell.

^{1}2008.

Relative transport costs $f40\_transport\_costs(k)$ are calibrated using total agricultural transport costs taken from the GTAP 7 database (see McDougall^{2} et al. 1998 for a general description of the GTAP model structure). GTAP agricultural transport costs represent transport costs from one sector to another sector (e.g. from a farm to the mill). Based on GTAP we calculate sector to sector transport costs for agricultatural inputs and outputs. In MAgPIE we want to represent market to market transport costs. Since markets are the links between sectors, we assume that the sum of 50% of the agricultural input transport cost and 50% of the output transport costs represent the agricultural transport costs for a commodity.

Relative transport costs $f40\_transport\_costs(k)$ are calculated by dividing total agricultural transport costs from GTAP 7 by the product of an initial (1995) cellular MAgPIE production pattern and cellular travel time. Total agricultural transport costs from GTAP 7 are based on total agricultural production in 2004. For consistency, we scale the GTAP data with the ratio of FAO production data for 2004 and 1995. Subsequently, MAgPIE is run several times in yield calibration mode until regional MAgPIE production is consistent with FAO production for the initial time step. Based on this first calibration run, total MAgPIE transportation costs are summed up and compared with GTAP data (see Figure 3). In case of low agreement, the calculation of relative transportation costs is repeated based on the calibrated MAgPIE production pattern and a second round of MAgPIE yield calibration is started. This process is repeated until MAgPIE total transport costs are in good agreement with GTAP total transport costs.

Limitations

The information in distances between production sites and markets is static over time, meaning that infrastructure is assumed to be static over time. Furthermore, transportation costs for non-ruminant products are excluded because currently there is no proper spatial allocation process for these products implemented.

### (B) off¶

In this realization the transport of goods is assumed to be free of charge. Correspondingly $vm\_cost\_transp$ is fixed to 0.

Limitations

No simulation of transportation (transportation free of charge).

## Definitions¶

Name Description Unit A B $f40\_distance(j)$ Transport distance to urban center minutes x $f40\_transport\_costs(k)$ Transport costs 2004US\$ / (ton DM * minute) x

_{The last columns of the table indicate the usage in the different realizations (numbered with capital letters)}

## Developer(s)¶

Jan Philipp Dietrich, Florian Humpenöder, Benjamin Bodirsky, Isabelle Weindl, Michael Krause

## See Also¶

core, 21_trade, 38_factor_costs, 39_landconversion, 90_presolve, overview

## References¶

^{1} Nelson, A. (2008) Estimated travel time to the nearest city of 50,000 or more people in year 2000. Global Environment Monitoring Unit - Joint Research Centre of the European Commission, Ispra Italy. Available at http://bioval.jrc.ec.europa.eu/products/gam/index.htm

^{2} McDougall, R.A., Elbehri, A., Truong, T.P., 1998. Global Trade Assistance and Protection: The GTAP 4 Data Base. Center for Global Trade Analysis, Purdue University.