05. Module III (Agro‐climatic constraints) - un-fao/gaezv5 GitHub Wiki
Introduction
When computing potential biomass and yields in Module II, initially no account is taken of the climate related impacts affecting production potential through pests and diseases, and unfavorable working conditions in the field. Such effects need to be included to arrive at realistic estimates of agro-climatic potential crop yields. Precise estimates of these impacts are very difficult to obtain for a global study. Here it has been approximated by quantifying the constraints in terms of reduction ratings, according to different types of constraints and their severity for each crop/LUT. Ratings vary by moisture regime, temperature regime and by level of inputs/management. The latter subdivision is necessary to take account of the fact that some constraints, such as bollworm on cotton, are present under low input conditions but are controllable under high input conditions in certain moisture regimes. While some constraints are common to all input levels, others (e.g., poor workability because of excess moisture) are more likely to affect operations under high input assumptions with fully mechanized cultivation. The main purpose of Module III is to evaluate crop growing conditions for possible agro-climatic constraints and to determine a respective yield reduction factor.
Agro-climatic constraints cause direct or indirect losses in the yield and quality of produce. Yields losses in a rain-fed crop due to agro-climatic constraints have been formulated based on principles and procedures originally proposed in FAO (1978) and successively expanded and updated from specialized literature, field data and CABI - Distribution Maps of Plant Pest and Diseases. Agro-climatic constraint updates were implemented repeatedly, e.g. FAO (FAO, 1980); FAO/UNDP (1982); Brammer et al. (1988); Kassam et al. (1991); UNDP/SSTC/FAO/SLA (1994); EISD/SRI (1999); FAO/IIASA (2000); Fischer, G., H. van Velthuizen (2002); FAO/IIASA (2012) and WWF/IIASA (2018). Four different yield constraints (i.e., yield-reducing factors) are accounted for:
- Pests, diseases, and weeds damage on plant growth (‘b’ group);
- Pests, diseases, and weeds damage on quality of produce (‘c’ group);
- Climatic factors affecting the efficiency of farming operations (‘d’ group), and
- Frost hazards (‘e’ group).
Although the constraints of group ‘d’ are not direct yield losses, such constraints do mean, for example, that the high input level mechanized cultivator cannot get onto the land to carry out operations. In practice, such limitations operate like yield reductions. Similarly, for the low input cultivator, for example, excessive wetness could mean that the produce is too wet to handle and remove, and again losses would be incurred even though the produce may be standing in the field.
Also included in this group, are constraints due to the cultivator having to use longer duration cultivars to enable harvesting in dry conditions. The use of such cultivars may incur yield restrictions, and such circumstances under wet conditions have therefore been incorporated in the severity ratings of agro-climatic constraints in group ‘d’.
The relationships between the occurrence of these constraints and quantified agro-climatic conditions, such as moisture stress and excess air humidity, and risk of early or late frost, are varying by location, between agricultural activities as well as by use of control measures. It has therefore been attempted to approximate the impact of these yield constraints on the basis of location-specific climatic conditions. The efficacy of control of these constraints (e.g. pest management) is accounted for through varying impact factors by levels of inputs/management.
Figure: Information flows of Module III
Still, there is relatively high level of uncertainty and therefore this quantification of agro-climatic constraints has been applied separately in Module III, such that effects are transparent, well separated and GAEZ assessments can be done with and without these constraints. This makes it also easy to apply alternative correction factors in studies where additional information on pests and diseases has been documented. Figure 5-1 gives an overview of the information flow in Module III.
In general, with increasing length of growing period and wetness, constraints due to pests and diseases (groups ‘b’ and ‘c’) become increasingly severe particularly to low input cultivators. As the length of growing period gets very long, even the high input level cultivator cannot always keep these constraints under control and they become severe yield reducing factors at all three levels of inputs. Other factors, such as poor pod set in soybean or poor quality in short lengths of growing period zones, are of similar severity for all three levels of inputs. Difficulties in lifting root crops under dry soil conditions (short lengths of growing periods group ‘d’) are rated more severely under the high level of inputs (mechanized) than under intermediate and low level of inputs. Agro-climatic constraints thus aim to represent any such additional direct or indirect losses of the yield and in the quality of produce. An explanation of the main yield-reducing components addressed by agro-climatic constraints is provided in the following sections.
Conceptual basis of agro-climatic constraint factors
The purpose of this section is to explain the conceptual basis of agro-climatic constraint factors considered in the model i.e., crop growth cycle and the length of growing period, water-stress during the growing period, pests, diseases, and weeds, climatic factors and frost hazard.
Mismatch between crop growth cycle and the length of the growing period
When the growing period is shorter than the growth cycle of the crop, from sowing to full maturity, there is loss of yield. The biomass and yield calculations account for direct losses by appropriately adjusting LAI and harvest index. However, the loss in the marketable value of the produce due to poor quality of the yield as influenced by incomplete yield formation (e.g., incomplete grain filling in grain crops resulting in shriveled grains or yield of a lower grade, incomplete bulking in root and tuber leading to a poor grade of ware), is not accounted for in the biomass and yield calculations. This loss is to be considered as an agro-climatic constraint in addition to the quantitative yield loss due to curtailment of the yield formation period. Yield losses can also occur when the length of the growing period is much longer than the length of the crop growth cycle, e.g. because of increased pest, disease and weed burden, excess wetness at harvest, or climatic conditions affecting the efficiency of farming operations.
Water-stress during the growing period
Water-stress generally affects crop growth, yield formation and quality of produce. The yield reducing impacts of water-stress vary from crop to crop. The total yield impact can be considered in terms of (i) the effect on growth of the whole crop, and (ii) the effect on yield formation and quality of produce. For some crops, the latter effect can be more severe than the former, particularly where the yield is a reproductive part (e.g., cereals) and yield formation depends on the sensitivity of floral parts and fruit set to water-stress (e.g., silk drying in maize).
Pests, diseases, and weeds
To assess the agro-climatic constraints of the pest, disease and weed complex, the effects on yields that operate through loss in crop growth potential (e.g., pest and diseases affecting vegetative parts in grain crops) are considered separately from effects on yield that operate directly on yield formation and quality of produce (e.g., cotton stainer affecting lint quality, grain mould in sorghum affecting both yield and grain quality).
Climatic factors directly or indirectly reducing yield and quality of produce
These include problems of poor seed set and/or maturity under cool or low temperature conditions, problems of seed germination in the panicle due to wet conditions at the end of grain filling, problems of poor quality lint due to wet conditions during the time of boll opening period in cotton, problems of poor seed set in wet conditions at the time of flowering in some grain crops, and problems of excessive vegetative growth and poor harvest index due to high night-time temperature or low diurnal range in temperature.
Climatic factors affecting the efficiency of farming operations and costs of production
Farming operations include those related to land preparation, sowing, cultivating and crop protection during crop growth, and harvesting (including operations related to handling the produce during harvest and the effectiveness of being able to dry the produce). Agro-climatic constraints in this category are expressed as workability constraints, which primarily account for excessive wetness conditions during necessary field operations. Limited workability can cause direct losses in yield and quality of produce, and/or impart a degree of relative unsuitability to an area for a given crop from the point of view of how effectively crop cultivation and produce handling can be conducted at a given level of inputs.
Frost hazard
The risk of occurrence of late and early frost increases substantially when mean temperatures drop below 10°C (Fischer, G., H. van Velthuizen, 2002). Hence, length of the thermal growing period with temperatures above 10°C (LGPt10) in a grid-cell has been compared with growth cycle length of frost sensitive crops. When the crop growth cycle is only slightly shorter than LGPt10 the constraints related to frost risk are adjudged moderate, when the growth cycle is very close or equal to LGPt10, the constraints have been adjudged as severe.
The availability of historical rainfall data has made it possible to derive the effect of rainfall variability through year-by-year calculation of yield losses due to water stress. Therefore the ‘a’ constraint, related to rainfall variability is no longer applied. Nevertheless, the ‘a‘ constraints have been retained in the agro-climatic constraints database for use with data sets containing only average rainfall data and for backward compatibility with earlier published AEZ information.
The ‘b’ and ‘d’ constraints and partly the ‘c’ constraint are closely related to wetness. The ratings of these constraints have been linked to indicators of wetness conditions, in Module I expressed by the number of growing period days (LGP) and/or as annual or seasonal moisture availability index P/ETo. While LGP may be curtailed in cooler climates by low temperatures despite of prevailing wetness, a high P/ETo ratio will capture conditions when precipitation tends to exceed evaporative demand and thereby indicate wetness. The ‘e’ constraint dealing with frost hazards is expressed in relation to the frost-free period LGPt10.
Box 5-1 Agro-climatic constraints context
In general, with increasing length of growing period and wetness, constraints due to pests and diseases (groups ‘b’ and ‘c’) become increasingly severe particularly to low input cultivators. As the length of growing period gets very long, even the high input level cultivator cannot keep these constraints under control and they become severe yield reducing factors at all three levels of inputs. Other factors, such as poor pod set in soybean or poor quality in short lengths of growing period zones, are of similar severity for all three levels of inputs. Difficulties in lifting root crops under dry soil conditions (short lengths of growing periods group ‘d’) are rated more severely under the high level of inputs (mechanized) than under intermediate and low level of inputs. For irrigated production the ‘c’ constraint is applied only at the wet end. Although the constraints of group ‘d’ are not direct yield losses in reality, such constraints do mean, for example, that the high input level mechanized cultivator, due to wetness, cannot get onto the land to carry out operations. In practice, this results in yield reductions. Similarly, for the low input cultivator, for example, excessive wetness could mean that the produce is too wet to handle and remove, and again losses would be incurred even though the produce may be standing in the field. Also included in this group are constraints due to the cultivator having to use longer duration cultivars to enable harvesting in dry conditions. The use of such cultivars incurs yield restrictions, and such circumstances under wet conditions have therefore been incorporated in the severity ratings of agro-climatic constraints in group ‘d’. --
In areas with year-round temperature growing periods, for example in the tropics and most of the sub-tropical thermal climate, the ‘b’, ‘c’ and ‘d’ agro-climatic constraints have been expressed for a big part in relation to LGP, in temperate and boreal climates equivalent LGP days are used as explanatory variable, which are calculated by an empirically estimated function of P/ETo ratios.
The wetness indicator used to interpolate damage factors from the look-up table (see table below) is based on both LGP and LGPeq, as follows:
$$ LGP_{agc} = \begin{cases} \min(120, \max(LGP, LGP_{eq})) & \text{if } LGP \leq 120 \ \max(210, \min(LGP, LGP_{eq})) & \text{if } 120 < LGP \leq 210 \ LGP & \text{if } LGP > 210 \end{cases} $$
The table below presents an example of agro-climatic constraints for rain-fed winter wheat. For irrigated production only the agro-climatic constraints related to excess wetness apply, as listed in the right half of the reduction factor table for LGPagc above 240 days. A listing of the agro-climatic constraint parameters considered for GAEZ crop/LUTs are presented in Appendix 5-1 Agro‐climatic constraints.
Table: Agro-climatic loss factors (in %) for rain-fed winter wheat, 40 days pre-dormancy +120 days post-dormancy
| LGP/LGPeq | 60–89 | 90–119 | 120–149 | 150–179 | 180–209 | 210–239 | 240–269 | 270–299 | 300–329 | 330–364 | 365 | 365+ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Low inputs | ||||||||||||
| b | 0 | 0 | 0 | 0 | 0 | 0 | 10 | 10 | 10 | 10 | 10 | 10 |
| c | 10 | 10 | 10 | 0 | 0 | 0 | 0 | 0 | 10 | 10 | 30 | 30 |
| d | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 10 | 30 | 30 |
| High inputs | ||||||||||||
| b | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 10 | 10 | 10 | 10 |
| c | 10 | 10 | 0 | 0 | 0 | 0 | 0 | 0 | 10 | 10 | 30 | 30 |
| d | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 10 | 10 | 10 | 30 | 30 |
| LGPt10 | 60–89 | 90–119 | 120–149 | 150–179 | 180–209 | 210–239 | 240–269 | 270–299 | 300–329 | 330–364 | 365 | |
| All input levels | ||||||||||||
| e | 100 | 50 | 25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
The ‘a’ constraint (yield losses due to rainfall variability) is not applied in the current assessment. This constraint has become redundant due to explicit quantification of yield variability through the application of year-by-year historical rainfall data sets.
Calculation procedures
The yield reduction factors for agro-climatic constraints were parameterized in lookup tables organized according to:
- Crop LUT;
- Thermal climate class;
- Number of actual/equivalent growing period days (LGP/LGPeq) for the ‘b’, ‘c’ and ‘d’ agro-climatic constraints;
- Length of the frost-free period (LGPt10) for the ‘e’ constraint, and
- Input level.
By combining the individual agro-climatic constraint factors ( fct _b,…, fct _e ) for constraint types ‘b’ to ‘e’, an overall yield reduction factor (fc3) is calculated for each LUT:
$$ fc_3 = \min \left( (1 - fct_b) \times (1 - fct_c) \times (1 - fct_d), 1 - fct_e \right) $$
With agro-climatic constraints evaluated, all three yield reduction factors (fc1 for thermal profile conditions and fc2 for soil moisture deficit calculated in Module II, fc3 for agro-climatic constraints calculated in Module III) are fully quantified and the agro-climatic potential crop yields are generated and mapped. Note that the evaluation of fc2 and fc3 is done separately for rain-fed and irrigated conditions. Factor fc1, though in principle the same for rain-fed and irrigated crops, can also vary by water source because crop calendars may differ between rain-fed and irrigated conditions and the selected defining LUT may differ as well.
See GAEZ v5 dataset on agro-climatic yield reduction factor (fc3) on the FAO Agro-Informatic Data Catalog (link).