Modeling climate change and impacts on crop production … · Modeling climate change and impacts...

University of Natural Resources and Life Sciences, Vienna Institute for Sustainable Economic Development

Modeling climate change and impacts on crop production in Austria

A Dissertation

by

Franziska Strauss

In partial fulfillment of the requirements for the degree

Doctor rerum naturalium technicarum

at the University of Natural Resources and Life Sciences, Vienna

Supervisor: Univ.Prof. DI Dr. Erwin Schmid Institute for Sustainable Economic Development,

BOKU, Austria Co-advisors: Dr. Elena Moltchanova Department of Mathematics and Statistics, UC, New Zealand Mag. Dr. Herbert Formayer Institute of Meteorology, BOKU, Austria DI Dr. Franz Sinabell Austrian Institute of Economic Research, WIFO, Austria

Vienna, May 2012

Table of Contents

Abstract ...................................................................................................................... ii

Kurzfassung.............................................................................................................. iii

1. Introduction............................................................................................................1

1.1 Research motivation and innovation ...............................................................1

1.2 Research objectives ........................................................................................4

2. Research methodology .........................................................................................5

2.1 Statistical climate modeling for Austria until 2040 ...........................................5

2.2 Integrated impact analyses of climate change ................................................7

3. Outline of the research articles and major findings ...........................................9

3.1 Modeling climate change and biophysical impacts of crop production in

the Austrian Marchfeld region .........................................................................10

3.2 High resolution climate data for Austria in the period from 2008 to 2040

from a statistical climate change model ..........................................................10

3.3 Modeling impacts of increased drought occurrences on crop production in

Austria ............................................................................................................11

3.4 How sensitive are estimates of carbon fixation in agricultural models to input data? ......................................................................................................12

3.5 Integrated assessment of crop management portfolios in adapting to climate change in the Marchfeld region...........................................................12

3.6 Optimal irrigation management strategies under weather uncertainty and

risk ..................................................................................................................13

3.7 Integrated assessment of large scale poplar plantations on croplands in Austria .............................................................................................................14

References ...............................................................................................................15

i

Abstract

Modeling impacts of climate change on crop production for the next three decades

requires climate change scenario data with a high degree of meteorological

consistency and spatial and temporal resolution. In the course of this cumulative

thesis, the statistical climate model ACLiReM (Austrian Climate model based on

Linear Regression Methods) has been developed to produce climate change

scenarios including daily weather data on minimum and maximum temperatures,

solar radiation, precipitation, relative humidity and wind speed for Austria at 1 km

grid resolution and the period 1975-2040. The climate change scenarios have been

statistically tested to assure physical and spatio-temporal consistencies. Developing

near future climate change scenarios by using statistical methods and observed

weather station data is seen as a valuable alternative to the well-known General

Circulation Models (GCMs) and Regional Climate Models (RCMs). Furthermore, the

statistical approach allows the modeling of extreme weather events such as

increased drought occurrences, which is also demonstrated in this thesis.

Biophysical process models like EPIC (Environmental Policy Integrated Climate)

have the potential to depict impacts on crop production of all anticipated variations in

input data (i.e. climate, topography, soils, management). Therefore, sensitivity

analyses have been performed which are important for providing valuable information

about the usefulness and appropriateness of such process models in impact studies

(e.g. for large scale applications) as well as for model inter-comparisons.

In this thesis, several economic optimization models have been developed and

applied in case study contexts. They integrate EPIC simulation data to derive optimal

crop management portfolios and investment strategies under consideration of

production risks and uncertainties both arising from future weather conditions. The

analyses of the thesis demonstrate that the developed climate change scenarios in

conjunction with EPIC and economic optimization models are adequate tools to

assess the impacts on climate sensitive sectors such as agriculture, which can be

used to design effective adaptation strategies.

ii

Kurzfassung

Die Modellierung von Klimafolgenwirkungen in der Landwirtschaft in den nchsten

drei Dekaden bedarf hoch aufgelster und konsistenter Klimawandeldaten. Im

Rahmen dieser kumulativen Dissertation wurde das statistische Klimamodell

ACLiReM (Austrian Climate model based on Linear Regression Methods) entwickelt,

um Klimawandelszenarien mit tglichen Wetterdaten zu Minimum- und

Maximumtemperatur, solarer Strahlung, Niederschlag, relativer Feuchtigkeit und

Windgeschwindigkeit zu generieren, welche fr sterreich auf einem 1 km

Rastergitter und in der Periode 1975-2040 verfgbar sind. Die Klimawandelszenarien

wurden auf ihre physikalische und rumlich-zeitliche Konsistenz statistisch geprft.

Der entwickelte statistische Ansatz unter Verwendung historischer Klimadaten ist

eine ntzliche Alternative zu den bekannten Generellen Zirkulationsmodellen (GCMs)

und Regionalen Klimamodellen (RCM) zur Generierung von Klimawandelszenarien

fr die nchsten zwei bis drei Dekaden. Darber hinaus erlaubt der statistische

Zugang die Modellierung von extremen Wetterereignissen (z.B. hufigeres Eintreten

von Drre), was auch in dieser Dissertation durchgefhrt wurde.

Biophysikalische Prozessmodelle wie EPIC (Environmental Policy Integrated

Climate) verfgen ber das Potential, die Auswirkungen aller antizipierten

Variationen von Inputdaten (Klima, Topographie, Boden, Bewirtschaftung) auf die

Pflanzenproduktion darzustellen. Folglich wurden Sensitivittsanalysen durchgefhrt,

welche hilfreiche Informationen ber die Ntzlichkeit und Eignung solcher

Prozessmodelle fr groflchige Anwendungen sowie fr Modellvergleichsstudien

liefern.

In der vorliegenden Dissertation wurden auch unterschiedliche konomische

Optimierungsmodelle entwickelt und im Rahmen von Fallstudien (z.B. fr die Region

Marchfeld) angewendet. Sie bauen auf den Simulationsergebnissen von EPIC auf.

Mit den Optimierungsmodellen wurden Portfolios von Pflanzenproduktionssystemen

und Investitionsstrategien fr Bewsserung unter Wetterunsicherheit und

Produktionsrisiko erstellt. Die Analysen in dieser Dissertation zeigen, dass die

entwickelten Klimawandelszenarien, EPIC und die konomischen

Optimierungsmodelle adquate Werkzeuge fr die Abschtzung der Auswirkungen in

klimasensiblen Sektoren wie der Landwirtschaft sind und fr die Analyse von

Adaptationsstrategien verwendet werden knnen.

iii

1. Introduction

1.1 Research motivation and innovation

Today it is nearly non-controversial that climate change associated with a global

temperature increase is proceeding. Nearly, because climate detractors exist who

may argue with a weak solar cycle and therefore with no temperature increase until

2035 (cp. Vahrenholt and Lning, 2012) or with uncertain effects of aerosols on the

climate. At Austrian scale, for instance, climate change is being observed and is

going to affect diverse economic domains in either positive or negative ways. In the

recent scientific literature, climate change impacts on agricultural production in

regions all over the world are being thoroughly investigated and widely discussed in

the scientific communities and the public (e.g. Alexandrow and Hoogenboom, 2000;

Schmid et al., 2004; Lobell et al., 2006; Lobell and Field, 2007; Tebaldi and Lobell,

2008; Balkovi et al., 2011). For instance, the Austrian Federal Ministry of

Agriculture, Forestry, Environment and Water Management (BMLFUW) financially

supports research for the establishment of a national adaptation strategy to climate

change in different sectors including agriculture, forestry, water, tourism and

electricity. The policy document recently published by BMLFUW (2010) provides first

recommendations for adaptation measures in these sectors.

For the near future until about 20401 , the well-known and commonly used

climate change scenarios from General Circulation Models (GCMs) and also

Regional Climate Models (RCMs) show their weaknesses due to (i) the low climate

change signal compared to the model internal variability on decadal time scale, (ii)

the random noise, and (iii) the time-lag between changing atmospheric conditions

and climatic impacts (Christensen et al., 2007; Randall et al., 2007; Cayan et al.,

2008). Moreover, a disadvantage of GCMs and RCMs is that temperatures and solar

radiation are often independently bias-corrected, and thus a reasonable correlation

with rainfall is not always guaranteed. Also the spatial resolutions are often not high

enough to show regional variations at site scale (Tebaldi and Sans, 2008). For all

these reasons, an alternative approach has been developed to produce near future

climate change scenarios for Austria, which is based on long-term historical daily

1 Throughout the thesis, the time span of interest is the period 20082040, denoted as the near future, which covers a typical climatological period (cp. definition of climate normals in a 30 year period from the World Meteorological OrganizationWMO).

1

weather data and statistical methods such as regression and bootstrapping

techniques. The statistical climate model ACLiReM (Austrian Climate model based

on Linear Regression Methods) has been developed and applied to produce a

spectrum of near future climate change scenarios for Austria, readily and publicly

available for further impact analyses (www.landnutzung.at). These climate change

scenarios are more consistent with respect to the physical interdependencies and the

spatio-temporal correlations between six weather parameters (i.e. minimum and

maximum temperatures, solar radiation, precipitation, relative humidity and wind

speed), and also represent well the local inter-annual variability and the small scale

climates in the complex topography of Austria. The near future climate change

scenarios also include deliberate assumptions on changing precipitation patterns,

because their future developments remain highly uncertain, such as (i) increases of

mean annual precipitation sums, (ii) decreases of mean annual precipitation sums,

and (iii) re-allocations in seasonal precipitation (i.e. increases in winter precipitation

and decreases in summer precipitation and vice versa). These assumed changes of

precipitation sums have been generated according to the suggestions in the literature

(IPCC, 2007; Jacob et al., 2008; Schnwiese, 2008; Gobiet et al., 2009; Eitzinger et

al., 2009; Auer et al., 2011). Furthermore, several climate change scenarios have

been developed, in which the frequency and intensity of future extreme weather

events have been manipulated to account for any possible increases of future

drought events.

The statistical approach developed in this thesis has limitations as well. One

essential limitation comes from the general assumption that the observed

temperature change continues linearly into the near future. This is only valid for the

next few decades and should not be used for any scenario development of the

second half of the century. Another limitation is that changes in the inter-annual

variability as suggested in several studies (e.g. Seneviratne et al., 2006) are not

considered. However, the magnitude of these changes is rather uncertain

(Christensen et al., 2007; Cayan et al., 2008). For the next three decades, it seems

more important to capture a realistic local inter-annual variability, which is indeed

provided by the statistical approach. However, variabilities on the decadal time scale

have not been included. Such variabilities have a magnitude of about 0.3 C for the

mean annual temperatures, and up to 10% for the mean annual precipitation sums

when comparing weather station data in Austria. This limitation can be partially

2

http:www.landnutzung.at

compensated by using ensembles of the developed climate change scenarios;

partially, because stochastic effects alone always lead to similar magnitudes for

mean annual temperatures, for example. However, deliberate assumptions for

scenario development can help to overcome limitations from neglecting annual and

decadal variations.

Knowing the advantages and disadvantages of the statistical climate change

scenarios, the second part of the thesis constitutes of various biophysical and

economic impact analyses of climate change on crop production in Austria. In

particular, the biophysical process model EPIC (Environmental Policy Integrated

Climate; Williams, 1995; Izaurralde et al., 2006) has been applied in all impact

analyses. Biophysical process simulation models are well suited for this task,

because all anticipated combinations and variations of input data (i.e. climate, soil,

topography, and crop management) can be simulated to depict the potential impact

spectrum on crop production and environment, on which also economic impact

analyses are often built (i.e. economic optimization models). EPIC is able to simulate

important natural processes in agricultural land management such as photosynthesis,

evapotranspiration, runoff, erosion, mineralization, nitrification and respiration. In this

context, EPIC has been applied to analyze its sensitivity to climate change and more

specifically to increased drought occurrences by applying climate datasets from

different sources. Such analyses are of particular interest and provide information

about the usefulness and appropriateness of deterministic process models for large

scale applications as well as for model inter-comparisons. In addition, EPIC has been

applied to provide biophysical data for several economic impact analyses. Economic

optimization models are well suited to help finding optimal land use and crop

management options as well as investment strategies in the context of production

risks and uncertainties due to changing climatic conditions. The integration of

biophysical data in economic optimization models allows a better representation of

the biophysical heterogeneity and interrelationships in spatial and temporal contexts.

In the course of this thesis, optimal crop management portfolios as well as optimal

investment strategies for irrigation under climate change have been detected for the

Marchfeld region in Austria. In addition, the biophysical and economic potentials of

large scale poplar plantations for bioenergy production have been analyzed on

Austrian croplands. These case study analyses should exemplify the capability of an

integrated modeling framework for climate change impact analyses in agriculture as

3

well as for developing cost-effective adaptation strategies (i.e. management and

investment) to better cope with the adverse effects of climate change.

1.2 Research objectives

One major objective of this thesis is the development of high resolution climate

change scenarios for Austria and the near future to be readily available for impact

analyses in climate sensitive sectors such as agriculture. These near future climate

change scenarios should be more consistent in a spatial and temporal context than

those usually provided by GCMs or RCMs. In addition, they should be physically

more consistent with respect to multiple weather parameters i.e. minimum and

maximum temperatures, precipitation, solar radiation, relative humidity and wind

speed, and thus shall be directly usable for impact analyses without the need of

employing complex bias-correction algorithms. Furthermore, the developed

methodology should allow a better quantification of uncertainties of near future

climatic conditions as well as the modeling of extreme weather events such as

droughts or floods in Austria.

Case study analyses in crop production should provide some information about

the appropriateness and usefulness of the developed climate change scenarios for

biophysical and economic impact analyses, which is another major objective of the

thesis. Modeling climate change impacts on agriculture require high spatial and

temporal resolutions of data input and model output to cope with the biophysical and

economic heterogeneity in Austrian crop production. Therefore, an integrated

modeling framework comprising detailed geo-referenced and statistical survey data

as well as the biophysical process model EPIC and economic optimization tools is

employed to conduct integrated impact analyses in case study contexts. In particular,

the case studies consist of analyses on (i) the sensitivity of EPIC to overall climate

change and extreme weather events as well as its performance compared to another

biophysical model, (ii) the development of an optimal crop management portfolio

under climate change for the Marchfeld region, (iii) the development of an investment

strategy for irrigation systems and an optimal irrigation portfolio under precipitation

uncertainty in the Marchfeld region, and (iv) the assessment of the economic

potentials and environmental consequences of large scale poplar plantations on

Austrian croplands.

4

2. Research methodology

2.1 Statistical climate modeling for Austria until 2040

In this thesis, regression techniques and repeated bootstrapping methods

together with historical daily weather data have been used to develop high resolution

near future (2008-2040) statistical climate change scenarios for Austria. The

developed climate change scenarios are based on different readily available datasets

mainly from the period 1975-2007 provided by the Central Institute for Meteorology

and Geodynamics (ZAMG). Clusters with homogenous climates in terms of mean

annual precipitation sums and temperatures (climate clusters) have been delineated

using the KLIM dataset with its 1 km spatial grid (Auer et al., 2000). In total, 60

climate clusters have been derived for Austria, and a representative weather station

has been selected for each climate cluster. Concerning temperature, a significant

trend has been estimated for Austria using the homogenized climate dataset

HISTALP (Auer et al., 2007), which includes monthly values of temperatures and

precipitation sums over the past period 1975-2007. In contrast, the annual

precipitation sums do not show any statistically significant development over this

period. Therefore, the developed statistical climate model ACLiReM has been applied

on the quality controlled daily time series of the StartClim project (cp. Schner et al.,

2003) solely for the parameters minimum and maximum temperatures by taking into

account the linear and seasonal time dependencies such as:

M

Y t (s) sin(2mt) (c) cos(2mt) (1) t m m t m1

where Y is the climate variable (maximum temperature and minimum

temperature), and t is the calendar time in years; sines and cosines represent

seasonal variability. The random residual is defined as the difference between the

daily temperature observations and the average historical temperature trend and

follows a Gaussian distribution. The stochastic part of the statistical climate model

comes from randomly selecting a month in the past to allocate the temperature

residuals. This bootstrapping procedure has been chosen because no significant

time dependence has been observed in the temperature residuals

5

(

(homoscedasticity). Moreover, the procedure of keeping monthly sequences

acknowledges the fact that typical weather events persist for a few days or longer.

Concerning the temperature development, a positive trend of 1.5 C has been

derived for Austria in the period 1975-2007 and extrapolated until 2040 (cp. also

Jacob et al., 2008).

The uncertainties in precipitation are captured by the bootstrapping method,

which has been applied in conjunction with quality controlled daily time series data of

precipitation (StartClim; Schner et al., 2003) to develop several precipitation

scenarios (e.g. from similar distribution of precipitation sums compared to the past to

increases or decreases of annual precipitation sums or re-allocation of seasonal

precipitation sums). The developed climate change scenarios also include data on

daily solar radiation, relative humidity and wind speed (all three parameters from the

corresponding original weather stations) to make them broadly applicable as well as

to better meet the specific data demands of typical biophysical impact simulation

models.

Climate change scenarios with increased severe weather have also been

developed for the period 2008-2040 by bootstrapping from a Beta-distribution of the

computed daily high-resolution weather dataset covering the period 1975-2007.

Therefore, precipitation patterns have been summarized in a drought index DId,

which is the fraction of the country area with zero precipitation on any given day:

2)

where (d) is the set of all climate clusters (60 climate clusters in Austria) dry

on day d, and Ac is the area of the climate cluster c ( corresponds to the total

area of Austria). The Beta-distribution approach can also be us

ed to develop other

extreme weather scenarios with e.g. increased occurrences of flood events.

In all statistical climate change scenarios, the temporal and spatial correlations

of solar radiation, relative humidity and wind speed with temperature and precipitation

data are by construction similar to the historical observations. This is an important

criterion of the statistical approach compared to GCMs or RCMs, particularly for near

future analyses.

6

2.2 Integrated impact analyses of climate change

The statistical climate change scenarios have been used for integrated impact

analyses of climate change on Austrian crop production in different case study

contexts. In particular, the climate change scenarios from ACLiReM together with

soil, topographical and crop management data have been input to the biophysical

process model EPIC. Geo-referenced topographical and soil data (BFW, 2009) have

been processed to delineate Homogenous Response Units for Austria (HRUs; cp.

Strmer et al., 2012), necessary to apply deterministic biophysical process models at

large scale. In total, 443 HRUs have been delineated for the Austrian agricultural

area and merged with the data of the 60 climate clusters. In addition, typical crop

rotation systems have been derived with the CropRota model (Schnhart et al.,

2011b) and typical crop management systems (i.e. fertilization, irrigation and tillage

operations) are in accordance with the guidelines of good agricultural practices

(Strmer et al., 2012).

EPIC simulates important biophysical processes in agricultural land use

management such as evapotranspiration, erosion, mineralization, nitrification and

respiration. Deterministic biophysical process models need sufficient input data in

good quality to reliably simulate outcomes of natural processes. If experimental data

are available, the models can be calibrated and the predictive accuracy can be

statistically tested. These models are also used to simulate impacts of climate

change on crop production and environment at larger scales e.g. from regional to

global scales (e.g. Balkovi at al., 2011; Schnhart et al., 2011a; Schneider et al.,

2011) by merging various sources of geo-referenced, statistical, survey, and

disciplinary data. However, the calibration of models becomes infeasible and model

validation is often restricted to comparisons between simulated and statistical crop

yields as well as to sensitivity analyses such as performed in this thesis. Therefore, a

detailed sensitivity analysis has been undertaken for the Marchfeld region which

demonstrates the correlation of EPIC outputs with gradually and discretely

manipulated weather parameters (discretized sensitivity). Another biophysical impact

analysis compares output data from two biophysical process models (EPIC and

Biosphere Energy Transfer Hydrology - BETHY/DLR) simulated with different input

data considering climate, land cover and land use data.

7

In order to assess the profitability of different crop management systems in the

Marchfeld region, a portfolio model has been developed which applies the

Conditional Value at Risk (CVaR) as a risk measure. One major advantage of the

CVaR compared to other risk measures is that it focuses on the tails of a distribution

and can therefore be applied for any distribution. The optimization problem

(maximization of expected crop production profits) has been performed under

different restrictions like different levels of risk aversion and the consideration of

additional environmental constraints (e.g. nitrate leaching and soil organic carbon

stocks which should not exceed/fall below a certain threshold). Thus, optimal crop

management portfolios under changing climatic conditions have been identified for

the Marchfeld region.

A stochastic dynamic programming approach has been developed to analyze

the probability and the optimal timing of investing into either a water-saving drip or a

sprinkler irrigation system in the Marchfeld region until 2040. Moreover, an optimal

irrigation management portfolio has been developed under consideration of different

degrees of risk aversion by using again the CVaR as a risk measure. The latter

approach allows investigating whether irrigation is part of an optimal production

portfolio and if so, which share of the production area is irrigated to minimize

production losses.

Finally, the Biomass optimization model Austria (BiomAT) has been used to

assess the economic potentials and environmental effects of large scale poplar

plantations on Austrian croplands, as short-rotation plantations seem to gain in

importance for producers of bioenergy. Biophysical properties of poplar plantations

from EPIC simulations and the corresponding gross margin annuities have been

input to BiomAT to assess the economic potentials (opportunity costs approach) as

well as environmental effects (i.e. total nitrogen emissions and topsoil organic carbon

stocks) of poplar production on croplands in Austria.

8

3. Outline of the research articles and major findings

All research articles presented in this thesis have been published in co

authorships. Therefore, the individual contributions are outlined, briefly. I have

initiated all first authored articles and provided an outline on the topic and methods to

conduct the analyses as well as the literature reviews and first drafts of the

manuscripts. In the course of the research process, these have been discussed,

reviewed and refined with the contributing co-authors. In particular, Herbert Formayer

has supported me in collecting most of the required climate datasets (especially

those from ZAMG; others have been available from online portals, for example).

Erwin Schmid and colleagues from the Institute for Sustainable Economic

Development have helped me in gathering the economic datasets. Erwin Schmid has

also performed the national simulations with EPIC and BiomAT. The prototype of the

statistical climate model for the Marchfeld has been developed in close collaboration

with Elena Moltchanova and Herbert Formayer. The first version of the statistical

climate model has been programmed together with Elena Moltchanova using the

statistical software package R. Supported by Herbert Formayer, I have prepared all

the datasets for the application of the statistical climate model (ACLiReM) to produce

climate clusters as well as the near future climate change scenarios for Austria. The

programming of the drought index has also been in close collaboration with Elena

Moltchanova and its impact analysis on crop production with Erwin Schmid. I have

performed most of the biophysical and economic impact analyses for the Marchfeld

region. The economic optimization models have been developed in close

collaboration with Sabine Fuss, Jana Szolgayov, Christine Heumesser, and Erwin

Schmid using GAMS (General Algebraic Modeling System). In the articles, which I

co-authored, I have provided data and simulation outputs as well as have contributed

to the analyses and manuscripts until publication.

9

3.1 Strauss F, Schmid E, Moltchanova E, Formayer H, Wang X. 2012. Modeling climate change and biophysical impacts of crop production in the

Austrian Marchfeld region, Climatic Change, 111: 641-664. DOI

10.1007/s10584-011-0171-0.

Research article 1 has been published in the journal Climatic Change. A

prototype of the statistical climate model has been developed for the Marchfeld

region. In addition, the sensitivity of selected EPIC outputs (crop yields, topsoil

organic carbon contents, and nitrate leaching) to input parameters such as minimum

and maximum temperatures, precipitation and solar radiation from different climate

dataset sources has been analyzed. The article contributes to the scientific literature

on modeling regional climate change by (i) providing an alternative statistical

approach to model climate change using historical daily weather data, (ii) testing the

sensitivity of the biophysical process model EPIC on alternative climate change

datasets, and (iii) comparing the statistical climate model outputs with GCM outputs

(i.e. scenarios from TYNDALL Centre for Climate Change Research: TYN SC 2.0, cp.

Mitchell et al., 2004). For the latter contribution, the emissions scenarios A1FI and B1

of the Parallel Climate Model (PCM; one of the GCMs) have been consulted at the

grid point nearest to the Marchfeld region. A trivial form of statistical downscaling has

been applied to make the climate change data comparable.

3.2 Strauss F, Formayer H, Schmid E. 2012. High resolution climate data for Austria in the period from 2008 to 2040 from a statistical climate change

model, International Journal of Climatology, DOI: 10.1002/joc.3434.

Research article 2 has been published in the International Journal of

Climatology. The main research objective of this article has been the development of

alternative climate change scenarios compared to the most often used climate

change projections of GCMs and RCMs, which are afflicted by rather important

limitations, predominantly for the next few decades. It builds on the prototype of the

statistical climate model presented in research article 1. This prototype has been

extended to the Austrian scale (ACLiReM) and applied to produce a spectrum of

climate change scenarios until 2040. This research article thus contributes to the

10

literature on producing high resolution climate change scenario data, which are

directly usable for impact analyses and publicly available (www.landnutzung.at). The

strength of this dataset is a proper consideration of the physical interdependencies

as well as the spatio-temporal correlations between the six weather parameters used

(minimum and maximum temperatures, solar radiation, precipitation, relative humidity

and wind speed). Moreover, it represents well the local inter-annual variability and the

small scale climates in the complex topography of Austria. It is especially valuable for

different impact studies at regional to national levels.

3.3 Strauss F, Moltchanova E, Schmid E. 2012. Modeling impacts of increased drought occurrences on crop production in Austria, Climatic

Change, submitted.

Research article 3 has been submitted to the journal Climatic Change. Based

on the climate change dataset developed in research article 2, hypothetical drought

scenarios have been generated for Austria. Therefore, a drought index has been

defined by using the empirical Beta-distribution as initial condition, which represents

the fraction of the country area with zero precipitation on any given day. Moreover,

the impacts of increased drought occurrences on crop yields and evapotranspiration

rates have been assessed with EPIC at national scale. The article thus contributes to

the literature in modeling increased drought occurrences at national scale as well as

the impacts on crop production. Only few articles in the scientific literature have

investigated the interrelations between increased drought occurrences and crop

production, yet. The impact analysis of increased drought occurrences on Austrian

crop production indicate that - for the areas with the past precipitation below the

annual average of ~850 mm - already little increases in dryness will result in

significantly lower crop yields (0.88% decrease in crop yield per 1% decrease in

precipitation).

11

http:www.landnutzung.at

3.4 Tum M, Strauss F, McCallum I, Gnther K, Schmid E. 2012. How sensitive are estimates of carbon fixation in agricultural models to input data?

Carbon Balance and Management, 7:3, DOI:10.1186/1750-0680-7-3.

In research article 4, the sensitivities of two biophysical process models, EPIC

and BETHY/DLR, have been assessed by comparing their simulation outputs using

various types of input data (climate, land cover and land use) for the Marchfeld

region. In particular, the developed statistical climate change dataset as well as

climate data from the European Centre for Medium-Range Weather Forecasts

(ECMWF) have been used in combination with land cover and land use data

(GLC2000 and the CORINE 2000 products). The article contributes to the literature in

presenting a detailed model inter-comparison study on the sensitivity of alternative

input data sources. Both biophysical models respond similarly to changes in input

data, albeit with a different magnitude. For single years, variabilities of up to 36% with

BETHY/DLR and of up to 39% with EPIC can occur in the net primary productivity

(NPP) under consideration of alternative input data. Moreover, the accuracy of land

cover and land use information has been investigated: GLC2000 land cover

classification overestimates the agricultural area of the Marchfeld region by 24%,

whereas the CORINE 2000 dataset overestimates land cover classification by only

7%. For case study analyses, the choice for land cover datasets providing more

regional contexts is thus recommended.

3.5 Strauss F, Fuss S, Szolgayov J, Schmid E. 2011. Integrated assessment of crop management portfolios in adapting to climate change in the

Marchfeld region, Jahrbuch der sterreichischen Gesellschaft fr

Agrarkonomie, Band 19(2): 11-20.

The impact of production risks and risk aversion on crop choices is a major

research question in agricultural economics. Research article 5 therefore contributes

to the literature in applying a portfolio optimization approach (based, for example, on

the Conditional Value at Risk - CVaR) to analyze the impact of climate change on

crop production risks and consequently on the choices of crops and crop

management. The source of risk comes from the developed stochastic climate

12

change scenarios. Inputs to the portfolio models are computed profit distributions

which differ among crops, crop managements and climate change scenarios. The

standard diversification effect can be shown such that the more risk averse a farmer,

the more diversification occurs. Another result is that minimum tillage always appears

in the optimal portfolios, however, with different management alternatives with

respect to straw harvesting and fertilization. In contrast, irrigation is not part of the

optimal portfolio, as the increasing revenues through higher crop yields cannot

compensate the higher production costs. However, the optimal portfolio results

strongly depend on the assumptions made with respect to irrigation management (i.e.

fixed irrigation amounts are taken into account) and the calculation of production

costs (e.g. level of water price in case of irrigation).

3.6 Strauss F, Heumesser C, Fuss S, Szolgayov J, Schmid E. 2011. Optimal irrigation management strategies under weather uncertainty and risk,

Jahrbuch der sterreichischen Gesellschaft fr Agrarkonomie, Band

20(2): 45-54.

Research article 6 contributes to the literature in assessing optimal irrigation

investment strategies considering weather uncertainty with two different model

approaches, the stochastic dynamic programming approach and the portfolio

optimization based again on CVaR as a risk measure. The static portfolio

optimization approach allows investigating whether irrigation is part of an optimal

production portfolio and if so, which share of the production area is irrigated to

minimize production losses. Optimal irrigation management portfolios have been

developed with respect to different degrees of risk aversion. On the other hand, the

stochastic dynamic programming approach examines the probability and the optimal

timing of investing in a water-efficient drip or a less water-efficient sprinkler irrigation

system until 2040. The stochastic dynamic programming approach shows a zero

probability for drip irrigation investment; the portfolio model also shows that drip

irrigation is never part of an optimal management portfolio under both risk neutrality

and risk aversion. However, sprinkler irrigation has a positive probability of being

adopted predominantly for the production of sugar beets and carrots. Future research

should be directed towards policy measures, e.g. implementation of water prices or

13

equipment subsidies which can increase the probability of adopting water-saving drip

irrigation systems (Heumesser et al., 2012).

3.7 Asamer V, Strmer B, Strauss F, Schmid E. 2011. Integrated assessment of large scale poplar plantations on croplands in Austria, Jahrbuch der

sterreichischen Gesellschaft fr Agrarkonomie, Band 19(2): 41-50.

Federal state governments and producers of bioenergy are planning a future

expansion of areas with short-rotation plantations. Research article 7 contributes to

the literature in assessing the economic potentials and environmental effects of large

scale poplar plantations on Austrian croplands in an integrative manner using the

models ACLiReM, EPIC and BiomAT. The latter allows determining the economic

potentials of biomass production on Austrian croplands in a spatially explicit way by

maximizing the gross margin annuities of agricultural land use under consideration of

available land, quality of location and respective crop rotations. The model results

show that the highest gross margin annuities are achieved in the south-eastern parts

of the country, however, the marginal opportunity costs would also be high compared

to other regions due to intensive livestock production. Thus, the lowest marginal

opportunity costs have been computed in the foothills of the Alps and in the north

eastern parts of Austria. Moreover, large scale poplar plantations would significantly

reduce total nitrogen emissions but only marginally increase topsoil organic carbon

stocks.

14

References

Alexandrow V, Hoogenboom G. 2000. Vulnerability and adaptation assessments of

agricultural crops under climate change in the Southeastern USA. Theoretical

and Applied Climatology 67:45-63.

Auer I, Bhm R, Mohnl H, Potzmann R, Schner W. 2000. KLIM - A digital

climatology of Austria 1961-1990. Proceedings of the 3rd European Conference

on Applied Climatology, 16 to 20 October 2000, Pisa, CD Rom, Institute of

Agrometeorology and Environmental Analysis, Florence.

Auer I, Bhm R, Jurkovic A, Lipa W, Orlik A, Potzmann R, Schner W, Ungersbck

M, Matulla C, Briffa K, Jones PD, Efthymiadis D, Brunetti M, Nanni T, Maugeri

M, Mercalli L, Mestre O, Moisselin J-M, Begert M, Mller-Westermeier G,

Kveton V, Bochnicek O, Stastny P, Lapin M, Szalai S, Szentimrey T, Cegnar T,

Dolinar M, Gajic-Capka M, Zaninovic K, Majstorovic Z, Nieplova E. 2007.

HISTALP - Historical instrumental climatological surface time series of the

greater Alpine region 1760-2003. International Journal of Climatology 27:17-46.

Auer I, Bhm R, Hofsttter M, Trk K, Kottek M. 2011. Long-term climate of

Carinthia: Historical climate trends, future scenarios and climate change indices

for the province of Carinthia. Final report of the Alpine Space Programme.

Balkovi J, Schmid E, Skalsk R, Novkov M. 2011. Modelling soil organic carbon

changes on arable land under climate change - a case study analysis of the

Kon Farm in Slovakia. Soil and Water Research 6:30-42.

BFW. 2009. Digital Soil Map for Austria. BFW, Vienna

(http://gis.lebensministerium.at/eBOD).

BMLFUW. 2010. Auf dem Weg zu einer nationalen Anpassungsstrategie, Policy

paper.

Cayan DR, Maurer EP, Dettinger MD, Tyree M, Hayhoe K. 2008. Climate change

scenarios for the California region. Climatic Change 87(Suppl 1):S21-S42.

Christensen JH, Hewitson B, Busuioc A, Chen A, Gao X, Held I, Jones R, Kolli RK,

Kwon WT, Laprise R, Magaa Rueda V, Mearns L, Menndez CG, Risnen J,

Rinke A, Sarr A, Whetton P. 2007. Regional climate projections. In: Solomon S,

Qin D, Manning M, Chen Z, Marquis M, Averyt KB, TignorM, Miller HL (eds)

Climate change 2007: the physical science basis. Contribution of working group

15

http://gis.lebensministerium.at/eBOD

I to the fourth assessment report of the intergovernmental panel on climate

change. Cambridge University Press, Cambridge.

Eitzinger J, Kersebaum Ch, Formayer H. 2009. Landwirtschaft im Klimawandel -

Auswirkungen und Anpassungsstrategien fr die Land- und Forstwirtschaft in

Mitteleuropa, Agrimedia GmbH.

Gobiet A, Heinrich G, Steiner M, Leuprecht A, Themel M, Schaumberger A,

Buchgraber K. 2009. AgroClim2 - Landwirtschaftliche Ertragsentwicklung und

Trockengefhrdung unter genderten Klimabedingungen in der Steiermark,

Endbericht.

Heumesser C, Fuss S, Szolgayov J, Strauss F, Schmid E. 2012. Investment in

irrigation systems under precipitation uncertainty. Submitted to Water

Resources Management [1st revision].

IPCC. 2007. Climate Change 2007 - The Physical Science Basis. Contribution of

Working Group I to the Fourth Assessment Report of the IPCC. ISBN 978 0521

88009-1.

Izaurralde RC, Williams JR, McGill WB, Rosenberg NJ, Quiroga MC. 2006.

Simulating soil C dynamics with EPIC: Model description and testing against

long-term data. Ecological Modelling 192(3-4):362-384.

Jacob D, Gttel H, Kotlarski S, Lorenz P, Sieck K. 2008. Klimaauswirkungen und

Anpassung in Deutschland - Phase 1: Erstellung regionaler Klimaszenarien fr

Deutschland. Abschlussbericht zum UFOPLAN-Vorhaben 20441138, Max

Planck-Institut fr Meteorologie (MPI-M), Hamburg.

Lobell DB, Field CB. 2007. Global scale climate-crop yield relationships and the

impacts of recent warming. Environmental Research Letters. doi:10.1088/1748

9326/2/1/014002.

Lobell DB, Field CB, Cahill KN, Bonfils C. 2006. Impacts of future climate change on

California perennial crop yields: model projections with climate and crop

uncertainties. Agricultural and Forest Meteorology 141:208-218.

Mitchell TD, Carter TR, Jones PD, Hulme M, New M. 2004. A comprehensive set of

high-resolution grids of monthly climate for Europe and the globe: the observed

record (1901-2000) and 16 scenarios (2001-2100). Tyndall Centre, Working

Paper 55, Norwich, UK.

Randall DA, Wood RA, Bony S, Colman R, Fichefet T, Fyfe J, Kattsov V, Pitman A,

Shukla J, Srinivasan J, Stouffer RJ, Sumi A, Taylor KE. 2007. Cilmate models

16

and their evaluation. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M,

Averyt KB, Tignor M, Miller HL (eds) Climate change 2007: the physical science

basis. Contribution of working group I to the fourth assessment report of the

intergovernmental panel on climate change. Cambridge University Press,

Cambridge.

Schmid E, Sinabell F, Liebhard P. 2004. Effects of reduced tillage systems and cover

crops on sugar beet yield and quality, ground water recharge and nitrogen

leaching in the Pannonic Region Marchfeld, Austria.

Pflanzenbauwissenschaften 8(1):1-9.

Schneider UA, Havlik P, Schmid E, Valin H, Mosnier A, Obersteiner M, Bottcher H,

Skalsky R, Balkovi J, Sauer T, Fritz S. 2011. Impacts of population growth,

economic development, and technical change on global food production and

consumption. Agricultural Systems 104(2):204-215.

Schner W, Auer I, Bhm R, Thaler S. 2003. Qualittskontrolle und statistische

Eigenschaften ausgewhlter Klimaparameter auf Tageswertbasis im Hinblick

auf Extremwerteanalysen, Teilprojekt von StartClim Startprojekt Klimaschutz:

Erste Analysen extremer Wetterereignisse und ihrer Auswirkungen in

sterreich, Wien.

Schnhart M, Schauppenlehner T, Schmid E, Muhar A. 2011a. Integration of bio

physical and economic models to analyze management intensity and landscape

structure effects at farm and landscape level. Agricultural Systems 104(2):122

134.

Schnhart M, Schmid E, Schneider U. 2011b. Crop Rota - A crop rotation model to

support integrated land use assessments. European Journal of Agronomy

34(4):263-277.

Schnwiese CD. 2008. Meteorologischer Hintergrund III - Extremereignisse aus

meteorologisch-statistischer Sicht. Promet (Deutscher Wetterdienst), Jahrg. 34,

1-2, 61-65.

Seneviratne SI, Lthi D, Litschi M, Schr C. 2006. Land-atmosphere coupling and

climate change in Europe. Nature 443:205-209.

Strmer B, Schmidt J, Schmid E, Sinabell F. 2012. Implications of agricultural

bioenergy crop production in a land constrained economy - the example of

Austria. Land Use Policy [accepted].

17

Tebaldi C, Lobell DB. 2008. Towards probabilistic projections of climate change

impacts on global crop yields. Geophysical Research Letters.

doi:10.1029/2008GL033423.

Tebaldi C, Sans B. 2008. Joint projections of temperature and precipitation change

from multiple climate models: a hierarchical Bayesian approach. Journal of the

Royal Statistical Society A 172(1).

Vahrenholt F., Lning S. 2012. Die kalte Sonne, Warum die Klimakatastrophe nicht

stattfindet, Hoffmann und Campe, 445 pp, ISBN-10: 3455502504.

Williams JR. 1995. The EPIC Model. In: Singh, V.P. (eds.). Computer Models of

Watershed Hydrology, Water Resources Publications, Highlands Ranch,

Colorado, 909-1000.

18

Both Research article 1 and Research article 2 are available at the respective

Journals webpage

Research article 3 is under review

Research article 4:

Tum M, Strauss F, McCallum I, Gnther K, Schmid E. 2012. How sensitive are

estimates of carbon fixation in agricultural models to input data? Carbon Balance and

Management, 7:3, DOI:10.1186/1750-0680-7-3.

Tum et al. Carbon Balance and Management 2012, 7:3 http://www.cbmjournal.com/content/7/1/3

R E S E A R C H Open Access

How sensitive are estimates of carbon fixation in agricultural models to input data? Markus Tum1,3*, Franziska Strauss2, Ian McCallum3, Kurt Gnther1 and Erwin Schmid2

Abstract

Background: Process based vegetation models are central to understand the hydrological and carbon cycle. To achieve useful results at regional to global scales, such models require various input data from a wide range of earth observations. Since the geographical extent of these datasets varies from local to global scale, data quality and validity is of major interest when they are chosen for use. It is important to assess the effect of different input datasets in terms of quality to model outputs. In this article, we reflect on both: the uncertainty in input data and the reliability of model results. For our case study analysis we selected the Marchfeld region in Austria. We used independent meteorological datasets from the Central Institute for Meteorology and Geodynamics and the European Centre for Medium-Range Weather Forecasts (ECMWF). Land cover / land use information was taken from the GLC2000 and the CORINE 2000 products.

Results: For our case study analysis we selected two different process based models: the Environmental Policy Integrated Climate (EPIC) and the Biosphere Energy Transfer Hydrology (BETHY/DLR) model. Both process models show a congruent pattern to changes in input data. The annual variability of NPP reaches 36% for BETHY/DLR and 39% for EPIC when changing major input datasets. However, EPIC is less sensitive to meteorological input data than BETHY/DLR. The ECMWF maximum temperatures show a systematic pattern. Temperatures above 20C are overestimated, whereas temperatures below 20C are underestimated, resulting in an overall underestimation of NPP in both models. Besides, BETHY/DLR is sensitive to the choice and accuracy of the land cover product.

Discussion: This study shows that the impact of input data uncertainty on modelling results need to be assessed: whenever the models are applied under new conditions, local data should be used for both input and result comparison.

Keywords: agricultural models, net primary productivity, EPIC, BETHY/DLR, land cover, weather

Background Modelling the net carbon uptake by vegetation (Net Primary Productivity, NPP) and estimating the yields of agricultural plants have become important tools to study the mechanisms of carbon exchange between the atmosphere and vegetation, as well as issues of food security. Different approaches are currently tracked which can be grouped to their approaches how photosynthesis is modelled. Models describing the chemical, physical and plant

physiological processes of plant development and the

* Correspondence: [email protected] 1Deutsches Zentrum fr Luft- und Raumfahrt (DLR), Deutsches Fernerkundungsdatenzentrum (DFD), Oberpfaffenhofen, D-82234 Wessling, Germany Full list of author information is available at the end of the article

interaction of plants with the atmosphere can be applied to simulate the rate of carbon dioxide uptake of the plant through photosynthesis (called Gross Primary Productivity, GPP). These models follow the concept of [1] and [2] to simulate the process of photosynthesis. Moreover, carbon uptake of well-watered and fertilized annual plants is linearly related to the amount of absorbed Photosynthetically Active Radiation (PAR), which can be derived from satellite data (i.e. the fraction of PAR which is absorbed by the canopy; cp. [3] or calculated by the accumulation of dry matter. NPP is defined as the difference between GPP and

autotrophic respiration. Therefore, it is important to estimate the autotrophic respiration of plants following the determination of GPP. Autotrophic respiration is defined as the oxidation of organic compounds found in

2012 Tum et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

mailto:[email protected]://creativecommons.org/licenses/by/2.0http://www.cbmjournal.com/content/7/1/3

Tum et al. Carbon Balance and Management 2012, 7:3 Page 2 of 13 http://www.cbmjournal.com/content/7/1/3

roots, stems and leaves, to CO2 or water. In the literature, different approaches to estimate autotrophic respiration are discussed, taking into account the actual biomass or GPP (e.g. [4-6]). When the Light Use Efficiency (LUE) approach is integrated in a coupled soil plant - atmosphere model as in the EPIC (Environment Policy Integrated Climate) model, daily estimates of evapotranspiration and carbon assimilation fluxes can be obtained [7]. In contrast to these models, more sophisticated

approaches are in use and under development. These models track photosynthesis on the molecule level. They take into account the interaction between plants, atmosphere and soil by simulating the uptake and release of carbon by plants and soil in a physically consistent way including conservation of energy and momentum. In the literature one can find descriptions of estab

lished vegetation models for use on different scales [8-11]. Each of these models is driven by meteorological input data and parameterized for global use with special focus on the long-term competition between the plant functional types when natural disturbance and succession driven by light competition occur. Models with a spatial resolution of kilometres and a time horizon of some years as e.g. the soil-vegetation-atmosphere-transfer (SVAT) model BETHY/DLR (Biosphere Energy Transfer Hydrology Model) [12] which can be used for regional assessments of NPP or biomass development. During the last decades, the use of both modelling

approaches was often met with resistance, mainly because of the need of calibration, validation and determination of the level of uncertainty (e.g.: [13-15]). Furthermore for many users, i.e. policy makers, it is difficult to judge whether the model outputs are within acceptable levels of uncertainty or not, mainly due to their limited background in model development [16]. However, in this context it is of importance to the policy maker to understand the validity of the model results and their associated uncertainties. Since empirical research traditionally advances in its

data accuracy and validity - in contrast - process-based models do not always provide comparable outputs, it is difficult to judge on the quality of modelled data, especially with the traditional criteria for assessing scientific outcomes [17]. However, regardless of the datas source, there will always be some uncertainty associated with it. To address these issues, we have assessed the variabil

ity of the soil-vegetation-atmosphere-transfer model BETHY/DLR [12] and the bio-physical process model EPIC [7] on three different meteorological input datasets and two land cover maps. Since the two models were designed for different specific purposes, we do not intend to discuss advantages or disadvantages but place special attention on the investigation of model

sensitivity to the spatial resolution of the input datasets. The Austrian Marchfeld region has been chosen as case study analysis because many datasets (Table 1) are readily available. The period of investigation is 2000 to 2003. It is important to note that this study is not a classical sensitivity analysis for assessing systematically the responses of models to changes in input data and model parameters (e.g. [18-21]), but a model variability analysis.

Methods Biophysical process models EPIC is a comprehensive model under continuous development since 1981, capable of simulating many agricultural processes that occur as a result of climate forcing, landscape characteristics, soil conditions and crop management schemes [7,22,23]. Biophysical processes simulated with EPIC include among others plant and crop growth, hydrology, wind and water erosion, and nutrient cycling. These processes are simulated with daily time steps or smaller. EPIC contains algorithms that allow for a complete description of the hydrological balance at the small watershed scale (up to 100 ha) including snowmelt, surface runoff, infiltration, soil water content, percolation, lateral flow, water table dynamics, and evapotranspiration. Daily weather can be endogenously generated for precipitation, temperature, solar radiation, wind speed, and relative humidity or it can be input exogenously. EPIC uses the concept of radiation-use efficiency by

which a fraction of daily photosynthetically active radiation is intercepted by the plant canopy and converted into plant biomass. The leaf area index is simulated as a function of heat units, crop stress and development stages. Daily gains in plant biomass are affected by vapor pressure deficits and atmospheric CO2 concentration [24]. Crop yield is simulated using the harvest index which is affected by the heat unit factor and includes the amount of the crop removed from the field as well as the above-ground biomass. Stress indices for water, temperature, nitrogen, phosphorus and aeration are calculated daily using the value of the most severe of these stresses to reduce potential plant growth and crop yield. Similarly, stress factors for soil strength, temperature, and aluminum toxicity are used to adjust potential root growth [25]. The soil water balance depending on the potential

water use, the root zone depth and the water use distribution parameter is applied in a general water use function where any water deficit can be overcome if a layer that is encountered has adequate water storage. The potential water use is reduced when the soil water storage is less than 25% of plant-available soil water by using dependencies on the soil water contents at field capacity and wilting point [7].

http://www.cbmjournal.com/content/7/1/3


Table 1 Meteorological, land cover, and other data.

Datatype Period used Resolution of space and time

Parameters used Characteristics References

Meteorology 2000-2003 weather stations; daily

Precipitation; Minimum

Measured time series Central Institute for Meteorology and Geodynamics (ZAMG)

temperature; Maximum temperature; Wind Speed; Radiation; Relative Humidity

Meteorology 2000-2003 1 km2 grid; daily Precipitation; Minimum

Reallocated time series (from now on ZAMG reallocated)

[32]

temperature; Maximum temperature; Wind Speed; Radiation; Relative Humidity

Meteorology 2000-2003 0.25; up to 4 times a day

Precipitation; Minimum

Time series of model re-analysis (ERA-40)

European Centre for Medium-Range Weather Forecasts (ECMWF)

temperature; Maximum temperature; Wind Speed; Cloud cover; Soil Water Content

Vegetation Indices

2000-2003 1 km2 grid; 36 time steps per year

Leaf Area Index (LAI)

Global coverage Ple dObservation des Surfaces continentales par TELdtection (POSTEL)

Landcover 2000 1 km2 grid, year 2000

Land cover / land use information;

Global coverage (GLC2000) [35,36]

22 classes

Landcover 2000 1 ha grid, year 2000

Land cover / land use information; 44 classes

European coverage (CORINE 2000)

[37]

Census 1999 Agrarstrukturerhebung

Marchfeld subregions, year 1999

Agricultural land use information; Main soil type distribution

Land use data of farms aggregated to municipalities

[30,42]

BETHY/DLR belongs to the family of SVAT models, which track the transformation of atmospheric carbon dioxide into energy storing sugars, a process known as photosynthesis. BETHY/DLR is based on the Jena Scheme of Atmosphere Biosphere Coupling in Hamburg (JSBACH) by [4] and was modified by [12]. The JSBACH model was originally considered for global usage and computes the biosphere-atmosphere exchange within the Global Circulation Model ECHAM5 (European Centre Hamburg). BETHY/DLR as well as JSBACH use the combined approach to integrate photosynthesis [26,27], which means that the enzyme kinetics are parameterized on the leaf level. In this context, C3 and C4 plants are distinguished because of significant differences in the way of their carbon-fixation: C4 plants (e.g. corn and sugar cane) are able to fix more atmospheric carbon dioxide at high temperatures than C3 plants (e.g. wheat and barley). Thus, the photosynthesis of C3 plants is saturated at higher temperatures. In a second step, the rate of photosynthesis is extrapolated

from leaf to canopy level by taking into account both, the canopy structure as well as the interaction of the plant between soil, atmosphere and vegetation. The two-flux scheme of [28] which includes three canopy layers, is used to approximate the radiation absorption in the canopy. Evapotranspiration, stomatal conductance and the soil water balance is included in the model formulation. To compute NPP on an annual basis snow is included in the water budget. Water stress is considered by calculating the demand for evapotranspiration using the approach of [2] limited by the criteria of [29]. Here it is assumed, that evapotranspiration can not be higher than a certain soil water supply via roots. Autotrophic respiration is evaluated as the sum of maintenance and growth respiration. The plant specific dark respiration determines the maintenance respiration, while growth respiration is assumed to be proportional to the difference between GPP and maintenance respiration. The main outputs of BETHY/DLR are given by time series of GPP, NPP, evapotranspiration, and of soil water



content in daily steps with the spatial resolution of the respective land cover classification. A more detailed model description can be found in [12]. The general characteristics as e.g. main outputs and

the general formulation to compute NPP of the two models BETHY/DLR and EPIC are presented in table 2.

Framework of Case Study Analysis The Austrian Marchfeld region serves as case study area to assess the variability of the two biophysical process models on alternative input datasets. The EPIC model has already been applied and validated here [30], and the data necessary for our study is readily available (see table 1). The Marchfeld region is located in Lower Austria, part of the Vienna Basin, and forms with around 100,000 ha one of the largest plains in Austria. Around 75% of the area is used for agricultural production. The natural boundaries are to the East the river March (the Austrian border to Slovakia), to the North the hills of the Weinviertel, to the West the mountain range of Bisamberg and the city of Vienna, and to the South the river Danube. For locating the region a map is presented in Figure 1. Since land use practices are not homogenously distrib

uted in this area, five sub-regions have been identified using the cluster analysis methods [31]. Each sub-region has an area of in between 85 km2 and 250 km2 . The urban land cover as well as forest and shrub lands have not been taken into account in the variability analysis. Five typical soils have been selected with respect to majority criteria for the agricultural land cover (four different Chernozems and one black earth; [30]). The biophysical process models have been applied

with different meteorological inputs (table 1) from the period 2000 to 2003. We have used meteorological observations from weather stations of the Central Institute for Meteorology and Geodynamics (ZAMG) in the Marchfeld region, reallocated meteorological data from weather stations across Austria of ZAMG [32], and meteorological data from the European Centre for Medium-Range Weather Forecasts (ECMWF).

Figure 1 Map of the study area. The case study area Marchfeld with the four sub-regions (upper figure), with underlying CORINE land cover dataset 2000. Green pixels represent forest, red and violet pixels urban areas, brown pixels shrub land, and yellow pixels agricultural areas. The lower figure highlights the location of the Marchfeld region. The red square represents the map extract of the upper figure.

The meteorological observations (ZAMG) are from the weather station in Gross Enzersdorf, and provide daily values of six weather parameters including minimum and maximum temperatures, relative humidity, wind speed precipitation and solar radiation.

Table 2 General characteristics of the biophysical process models EPIC and BETHY/DLR.

BETHY/DLR EPIC

Abbreviation Biosphere Energy Transfer Hydrology Environmental Policy Integrated Climate Model

References [4,12] [7,22,23]

Model type SVAT model crop model

Time step Daily Up to < 1 day

Main simulation processes GPP, NPP, NEP, evapotranspiration, soil plant and crop growth, heat and water balance, wind and water water content erosion, nutrient cycling

General formulation to NPP = GPP - autotrophic respiration NPP = (yield+straw+roots)- (water content+non carbon fraction) compute NPP



[32] developed a reallocated meteorological dataset comprising climate data for Austria and the period from 1975 to 2007 with temporal and spatial resolutions of one day and 1 km2 . In addition climate change scenarios have been developed for the period 2008 to 2040. They processed daily data from 34 weather stations of ZAMG to 60 spatial climate clusters with homogeneous climates relating to mean annual precipitation sums and mean annual temperatures from the period 1961-1990. Based on these precipitation and temperature classes four climate clusters describe the climate in the Marchfeld region (cluster 1: mean annual precipitation sums smaller than 500 mm and mean annual temperatures between 8.5C and 9.5C; cluster 2: mean annual precipitation sums smaller than 500 mm and mean annual temperatures between 9.5C and 10.5C; cluster 3: mean annual precipitation sums between 500 mm and 600 mm and mean annual temperatures between 8.5C and 9.5C; cluster 4: mean annual precipitation sums between 500 mm and 600 mm and mean annual temperatures between 9.5C and 10.5C). For each homogenous climate cluster, [32] performed regression model analyses primarily to compute a set of daily climate data for the time period 2008 to 2040. This method has also been applied for the time period 1975 to 2007 to provide a consistent dataset. The integral parts of the regression model are i) the consideration (extrapolation in the period 2008 to 2040, respectively) of the observed linear temperature trend from 1975 to 2007 derived from a homogenized dataset, and ii) the repeated bootstrapping of temperature residuals and of observations for solar radiation, precipitation, relative humidity, and wind speed to ensure consistent spatial and temporal correlations. We have also used these reallocated data for the period 2000 to 2003 in our variability analysis. The third dataset is derived from ECMWF data and

has a temporal resolution of up to four times a day and a spatial resolution of 0.25 0.25. It includes model analysis data of 2 m air temperature, cloud cover, soil water content of the four upper layers and wind speed at 10 m above ground. From this dataset the daily mean, as well as minimum and maximum temperatures and the daily mean of cloud cover in all three strata (high, medium, low) are used. The daily temperature values are scaled with the difference between ECMWF reference height and the global ETOP05 (Earth Topography and Ocean Bathymetry Database) 5-minute gridded elevation data by using the temperature gradient of the U.S. Standard Atmosphere (-0.65 K per 100 m) in order to downscale the ECMWF temperature data to km2 resolution. Precipitation values are derived twice a day from the ECMWF re-analysis project (ERA-40). PAR is not used directly from the corresponding ECMWF product data as it is only available as forecast

data and therefore rather uncertain. Thus, daily PAR is determined from global radiation which is computed following the approach of [33] taking into account the geographical coordinates of the day, and using a transmission, which depends on the degree of cloudiness. The degree of cloudiness is calculated as a weighted sum of each cloud strata for each day, and the global radiation is calculated for each location in the time step of one hour. The advantage of this approach is the use of analysis data of cloud coverage to compute PAR data which leads to more exact results than directly using the PAR forecast data [12]. Hence the BETHY/DLR model needs an initial soil

water content, the ECMWF soil water dataset is used only for the transient phase of the model. Afterwards the model simulates the soil water content independently, according to the hydrological boundary conditions. Investigations of [12] have shown that in most cases sufficient hydrological boundary conditions are available after a transient phase of about one year. In addition to the meteorological data, the BETHY/

DLR model is driven by two sets of remote sensing data. Detailed and homogenous land cover / land use information are used to get information about the vegetation types the model is run for. Vegetation is represented by time series of the Leaf Area Index (LAI). Time series of LAI were used from the Carbon cycle and Change in Land Observational Products from an Ensemble of Satellites (CYCLOPES) 10 day composite datasets of POSTEL (Pole dObservation des Surfaces continentales par TELedetection), which have a spatial resolution of 1 km 1 km. For each of the grid cells, time series analysis has been applied in order to eliminate data gaps and outliers. In the framework of this study the harmonic analysis has been used. The method of the harmonic analysis is based on the method of superposition such as the Fourier transformation. This method ([34]) is used to process LAI time series at the German Remote Sensing Data Center. The CYCLOPES dataset additionally contains informa

tion of land cover and land use and is available as GLC2000 (valid for the year 2000). The Land Cover Classification System of the Food and Agriculture Organization of the United Nations has been used to derive land cover classes of GLC2000 resulting in 22 different land cover classes [35,36]. A translation of the GLC2000 vegetation classes had

to be performed in order to use the GLC2000 land use / land cover classification to model NPP with BETHY/ DLR. The actual model setup of BETHY/DLR includes 33 inherent vegetation classes which can be regarded as vegetation types. Each vegetation type is linked with biochemical parameters as i.e. the maximum electron transport rate and the maximum carboxylation rate, and



other vegetation specific parameters as maximum height and rooting depth. These parameters describe the mechanism of photosynthesis of vegetation. In this study only the GLC2000 class 16 Cultivated and managed areas has been used and translated to the BETHY/ DLR vegetation type arable land as no further detailed information about the land use (e.g. crop rotation) is available from the GLC2000. In addition to the GLC2000 dataset the Coordinated

Information on the European Environment (CORINE) 2000 land cover / land use classification has been used, to validate the GLC2000 dataset. The CORINE 2000 data was derived from LANDSAT satellite images and is also available for the year 2000 [37]. The CORINE 2000 is available as raster datasets in spatial resolutions of 100 m 100 m, 250 m 250 m and 1 km 1 km for 32 European countries, including Austria. For this study the dataset with resolution 100 m 100 m has been used. The CORINE 2000 provides information about 44 vegetation classes which had also to be translated to BETHY/DLR vegetation types. We assumed that only the CORINE 2000 class Non-irrigated arable land contains the needed information about agricultural land, since all other classes which are available for the Marchfeld region report different land use (e.g. forests and urban areas). The CORINE 2000 class Non-irrigated arable land is then translated to the BETHY/DLR class arable land.

From Crop Yield to NPP The crop yields of EPIC for the thirteen crops in the Marchfeld region have been converted to NPP values (table 2) for comparison with the BETHY/DLR outputs, which are given as time series of NPP. For this purpose, conversion factors of the relation between yield and straw as well as the above- and below- ground biomass are used. Empirical conversion factors about the relations between crop yield and straw yield can be found in e.g. [38,39]. In a first step, the above-ground biomass is computed for each crop using these empirical conversion factors. In a second step the below-ground biomass is computed with the use of conversion factors about the ratio of above- to below- ground biomass which are described in [40]. These conversion factors which originally have been derived for crops in Canada are assumed to be valid for the area of interest as well, as it already was proposed by [41]. After calculating the biomass of the whole plant, the remaining water content and the non carbon content have to be subtracted, following crop specific values, which are also reported in e. g. [38]. A detailed description of the approach and the used factors can be found in [41]. In order to compare the now available NPP per crop

and sub-regions of EPIC with the BETHY/DLR results,

statistical data about the land use of each of the four sub-regions is used to aggregate the NPP of EPIC. These statistical data provided by [30] and [42] give detailed information about the distribution of agricultural area over the thirteen main crops as well as the distribution of the five main soils being representative for the Marchfeld region. The results of BETHY/DLR have been aggregated to annual sums per sub-region with a Geographic Information System (GIS) tool, taking into account the equi-rectangular projection (latitude longitude, WGS84 (World Geodetic System 1984)) of the data.

Results and discussion The variability analysis consists of seven model setups to compare model response to different input datasets. Three model simulations with the EPIC model have been performed and four with the BETHY/DLR model. The model setups are presented in table 3. The EPIC model requires homogeneity with respect to

data input (i.e. soil, topography, weather, crop management) such that the model has been applied for all combinations of climate, soil, and crop management, separately. Thus, the variability analysis has been conducted mainly for the meteorological datasets. In total 60 different model runs have been performed with EPIC for each crop. In contrast, the BETHY/DLR model is driven with the two different land cover classifications as well as the three different meteorological input data sets. For the Marchfeld region the FAO soil map of the world, which is used as input data for BETHY/DLR, reports one major soil type (Haplic Chermozem) which occupies 89% of the area and four additional soil types for the rest of the area. The EPIC model setup EPIC(1) is interpreted as reference, as it represents the already validated model setup [30]. In Figure 2, all model results fare compared to the

EPIC(1) results (table 3). The values of NPP are given in kilotonnes carbon per sub-region and year. Depending on the model setup, the NPP results of

BETHY/DLR show a variability of overestimations of up

Table 3 Model setups for the variability analysis

Model Meteorological Land cover Short input classification Name

BETHY/ ZAMG CORINE 2000 BETHY(1) DLR

ZAMG reallocated CORINE 2000 BETHY(2)

ECMWF CORINE 2000 BETHY(3)

ECMWF GLC2000 BETHY(4)

EPIC ZAMG - EPIC(1)

ZAMG reallocated - EPIC(2)

ECMWF - EPIC(3)



Figure 2 Comparison of the model results. Comparison of the model results (NPP) of BETHY/DLR and EPIC for the four Marchfeld sub-regions and the period 2000 to 2003. The nomenclature follows the scheme of table 3. Circles represent sub-region 1, triangles sub-region 2, crosses sub-region 3 and diamonds sub-region 4.

to 32% and underestimations of up to 12%, linked with modelled when using the GLC2000 and meteorological coefficients of determination between 0.94 and 0.63, input data from ECMWF (Figure 2D). Figure 2D reprerespectively. The highest overestimation of NPP (32%) is sents the results of both models with the typical setup



which was used in previous investigations (default setup). This overestimation is combined with a high coefficient of determination of about 0.94. When changing the land cover classification from GLC2000 to CORINE 2000 (while the meteorological input remains unchanged) an underestimation of about 12% has been found (Figure 2C). From Figure 2C it is evident that only 4 BETHY/DLR results determine the underestimation and thus the coefficient of determination of about 0.77. These four data points are all representative for sub-region 4, whereas the rest is close to the 45 line. Using measured meteorological data from ZAMG results in an overestimation of NPP of about 11% (Figure 2A), which is combined with the highest variability within the sub-regions and years for all four model setups of BETHY/DLR. Nevertheless a high coefficient of determination of about 0.68 is achieved. When using the reallocated ZAMG data of [32] for BETHY/DLR combined with CORINE as land cover an overestimation of the modelled NPP of about 15% (Figure 2B) has been found. A strong correlation of the simulation years is observed, which indicates homogeneity in the meteorological data. The comparison between EPIC results with different

weather input reveals that the ECMWF data affects the EPIC model to underestimate NPP by 8% (Figure 2F). The use of the reallocated meteorological dataset (Figure 2E) results in a little underestimation, linked with the highest coefficient of determination (0.97). Figure 2E demonstrates that EPIC is not very sensitive to measured or homogenized meteorological input data just in contrast to BETHY/DLR which can be seen in Figure 2A and 2B. Measured meteorological data during the four years result in a high variability of the annual NPP of sub-regions 1 and 4 while the reallocated meteorological data cluster the annual NPP of all sub-regions resulting in low variability for all sub-regions. Figure 2D and figure2F show that the EPIC model as

well as the BETHY/DLR model react in a similar way when alternating between ECMWF and ZAMG data. The BETHY/DLR model simulates 23% more NPP when using the ZAMG data, and the EPIC model simulates around 7% more NPP when using the ZAMG data. A reason for investigating the influence of different

land cover classifications (GLC2000 versus CORINE 2000) is the higher spatial resolution of CORINE 2000. It is expected that CORINE 2000 represents the small scale land use structure of the Marchfeld region better than the GLC2000 classification. In Figure 3 the agricultural areas reported in the statistical source [30,42], the GLC2000 and CORINE 2000 are presented for all four Marchfeld sub-regions. The agricultural areas presented in GLC2000 and

CORINE 2000 have been computed using GIS tools. As

Figure 3 Validation of Land cover land use products . Comparison of agricultural areas described by statistical sources [30,42], GLC2000 and CORINE 2000 in km2 for the four sub-regions of the Marchfeld region.

shown in Figure 3, the GLC2000 considerably overestimates the agricultural areas (sub-regions one, three and four) by 25% to 57% compared to the statistical information. On the other hand, CORINE 2000 slightly over(17%) or underestimates (6%) the agricultural areas compared to the statistical sources. However, approximately the same agricultural area is found for subregion two for each land cover classification. For all sub-regions of the Marchfeld region the statistical data report an agricultural area of around 670 km2 , GLC2000 of 881 km2 , and CORINE 2000 of 718 km2 . As the difference in agricultural area between CORINE 2000 and the statistical data is smaller than the difference between GLC2000 and the statistical data, we conclude that the CORINE 2000 land cover represents the real situation more precisely than GLC2000. The differences of the results described in Figure 2D and 2C showing an NPP decrease when changing from GLC2000 to CORINE 2000 can thus be explained by the fact that the BETHY/DLR model was driven for a smaller agricultural area. To proof this, the results for BETHY(3) and BETHY(4)

are presented in Figure 4 as a linear correlation. For both model setups meteorology was fix (ECMWF), but the land cover classification was changed. With this direct comparison it becomes clear that the reason for the highly different model results presented in Figure 2C and 2D lays in the uncertainty in the two land covers. When comparing the ECMWF data with the mea

sured ZAMG data it is obvious that the ECMWF data underestimates the maximum and minimum temperatures (see Figure 5). The comparison of daily weather measurements is conducted for two of the 34 ZAMG



Figure 4 Comparison of the BETHY/DLR model response to different land cover / land use products. Comparison of the model results (NPP) of the BETHY/DLR runs BETHY(3) and BETHY(4) for the four Marchfeld sub-regions and the period 2000 to 2003. The nomenclature follows the scheme of table 3. Circles represent sub-region 1, triangles sub-region 2, crosses sub-region 3 and diamonds sub-region 4.

weather stations which are situated closest to the Marchfeld (Schwechat and Gross Enzersdorf) and for the time period 2000 to 2003. For both stations the maximum temperature of the

ECMWF data is underestimated by about 21% expressed by a high coefficient of correlation of up to 0.72. The minimum temperature is underestimated even slightly higher (up to 28%) but again combined with a high coefficient of correlation (up to 0.74). In contrast, precipitation is not represented very well by ECMWF data as the correlation reveals high uncertainties. Hence a comparison of the ECMWF data for only two measurement stations is not very meaningful. Therefore the analysis has been expanded to all of the 34 available ZAMG weather stations. The analysis shows that the mean maximum and minimum temperatures of ECMWF data averaged over daily values in the period 2000 to 2003 are about 24% and 29% lower, respectively, than the temperatures recorded by the 34 ZAMG weather stations. However, minimum and maximum temperatures are both linked with a coefficient of determination of about 0.65, which is in good correspondence with the two presented observation stations in Figure 5. The comparison between sums of annual precipitation between the ECMWF and the ZAMG data reveals over- and underestimations of up to 90% for single stations. The daily precipitation rates averaged over all ZAMG observation stations show a coefficient of determination of about 0.27. This very low coefficient corresponds with the presented stations in Figure 5 and indicates poor agreement of measured and simulated precipitation.

As ECMWF data significantly underestimate temperature, the increase of NPP when using ZAMG data could be explained by longer vegetation periods in the ZAMG data. We investigated the vegetation period by computing the growing-degree-days (GDD). The basic equation is: GDD = [(TMAX + TMIN)/2]-TBase, where TMAX and TMIN are daily maximum and minimum temperatures, respectively and TBASE is the base temperature which can be fixed at 10C [43]. Furthermore, the growing period in Austria is assumed to be from mid March to mid October. The mean GDD averaged over all 34 ZAMG stations in Austria and the years 2000 to 2003 is about 1186.2, which is about 136.1 (~11.5%) more than the corresponding ECMWF GDD value (1050.1). In a third model setup both models are driven with

the reallocated ZAMG data to test the model response to homogenized trend data. Figure 2B and 2E show that both, the EPIC and the BETHY/DLR models respond in a consistent way, concerning their annual variability, to the reallocated ZAMG data. The variability in the NPP over the four years within one sub-region is about 4% (EPIC) and 3% (BETHY/DLR), respectively. To give information about the annual variability of

NPP within the m

Modeling climate change and impacts on crop production … · Modeling climate change and impacts...

Documents

Transcript of Modeling climate change and impacts on crop production … · Modeling climate change and impacts...