Matching soil grid unit resolution with polygon unit map scale is
important to minimize uncertainty of regional soil organic carbon
(SOC) pool simulation as their strong influences on the
uncertainty. A series of soil grid units at varying cell sizes were
derived from soil polygon units at the six map scales of
Soil organic carbon (SOC) is the largest terrestrial carbon pool
(Schlesinger, 1997), with stocks about four times the biotic
(trees, etc.) pool and about three times the atmospheric pool
(Lal, 2004). Relatively modest changes in SOC storage can result
in a significant alteration in the atmospheric
Agricultural soils are a highly sensitive part of the global
carbon cycle (Shi et al., 2010; Wang et al., 2011), carbon
sequestration by agricultural soils presents an immediate viable
option for increasing soil carbon pool and reducing atmospheric
The DNDC model developed by Li et al. (1992a, b) can simulate C and N biogeochemical cycles occurring in agricultural systems, driven by both the environmental factors (e.g. soil organic matter, texture, pH, bulk density, hydraulic properties, daily temperatures and precipitation, etc.) and management practices (e.g. crops, tillage, fertilization, manure application, grazing, etc.). It has been validated through long-term applications internationally at the plot scale, including many sites of North America, Europe, Asia, etc. (Pathak et al., 2005; Li et al., 2006; Tonitto et al., 2007), and is one of the most widely accepted biogeochemical models in the world (Li, 2007; Tang et al., 2006; Li et al., 2010).
The DNDC model has also been utilized to upscale estimates of SOC from plot to region scale. At the region scale the DNDC modelling conducted initially has used counties as basic simulation units, where minimum and maximum soil parameter values for each county were derived from soil maps to simulate an upper and a lower estimate of several C and N pools (Cai et al., 2003; Li et al., 2004). However, county scale model simulations are subject to great uncertainties as soil properties are averaged for each county, largely ignoring the nonlinear impacts of soil heterogeneity therein (Rüth and Lennartz, 2008; L. M. Zhang et al., 2014).
Recently for DNDC up-scaled utilization, a region is partitioned into many simulation units, within which all soil properties are assumed to be as homogeneous as they are at the plot scale (Li et al., 2005; Zhang et al., 2012). The homogeneity assumption is a possible major source of error when extending DNDC modelling from the plot to the region scale (Li et al., 2002, 2004). As the area of the basic simulation unit increases so does soil property variability or heterogeneity, calling into question the accuracy of its capture (Smith and Dobbie, 2001; Bouwman et al., 2002).
Soil polygons derived from soil vector maps are used as basic
simulation units, that is one way to reduce effects of soil
heterogeneity on DNDC modelling as they can as possible (Xu
et al., 2012b; Yu et al., 2013; Zhang et al., 2012). Even so, the
soil heterogeneity within a soil polygon unit still exists, and
depends on the soil vector map scale, smaller map scale resulting
higher heterogeneity (Yu et al., 2013). To different broad
regions, multi-scales of the polygon unit simulated with DNDC
ranged widely from
Another way to reduce effects of the soil heterogeneity on the DNDC modelling is that soil grid cells are used as the basic simulation units (Huang et al., 2004; Y. Q. Yu et al., 2007; Shi et al., 2010; Yu et al., 2011). Cell size or resolution of the soil grid units is one of rulers to scale the soil heterogeneity therein, lower resolution or larger cell size resulting higher soil heterogeneity likewise. The cell size or resolution take effects extremely also on the accuracy and uncertainly of the soil grid unit simulation with DNDC (Yu et al., 2011).
The soil grid units are more often applied to simulation of SOC
pool (Qiu et al., 2005; Tang et al., 2006; Yu et al., 2011; Liu
et al., 2011), as they are more easily manipulated for spatial
model simulation, geo-statistics and spatial analysis than the
soil polygon units (Huang et al., 2004; Li et al., 2005). They are
often derived by data conversion from the soil polygon units, but
the grid resolution choice varies by researcher even if the soil
polygon units are at same map scale and in same region (Y. Q. Yu et al.,
2007; Shi et al., 2010; Yu et al., 2012). For example, the soil
polygon units compiled in the Soil Database of China (Yu et al.,
2007a) at the map scale of
Our concerning is whether these soil grid units at different cell sizes are equivalent in accuracy or granularity to their parent soil polygon units at a corresponding map scale for DNDC modelling. In other words, whether these soil grid unit datasets regulate coarser data or contain redundant data of soil properties, contrasting to their parent soil polygon unit dataset at a certain map scale. The coarser or redundant dataset affects the simulation unit inner homogeneity of soil properties, and farther affects the common outcome too, being that modelling error will be lower if all features within the simulation unit are more homogeneous (Cai et al., 2003; Yu et al., 2011, 2013).
In fact the accuracy and the redundancy are two important issues to soil simulation units' dataset conversion from polygon to grid format, which are often neglected in modelling at regional scale. The accuracy of the grid unit dataset determine reliability and uncertainty of SOC grid simulation (Batjes, 2000; Ni, 2001), the redundancy of the dataset results in mistaken understanding of data accuracy and redundant workload and cost of the simulation (Yu et al., 2011, 2013). Some researches focus on data accuracy but neglect the data redundancy (Yu et al., 2007b; Shi et al., 2010), while others neglect the data accuracy (Batjes, 2000; Y. Q. Yu et al., 2007) when conduct data conversion, they always search for an individual solution in every case.
Given the variety of datasets and number of simulations, in combination with data accuracy and redundancy as well as computational costs (Schmidt et al., 2008), important questions are raised. How sensitive is DNDC modelling to different simulation units at varied vector map scales or raster grid resolutions? Which raster resolution is optimal to DNDC grid simulation at a fixed soil map scale for error and cost controls? Matching the soil grid unit resolution with polygon unit map scale is one of essential issues to DNDC modelling.
In the present study, paddy soil polygon simulation units at six
vector map scales from
The objectives of the study were to (1) reveal the impact of vector map scale and raster resolution of soil simulation units on the DNDC modelling, (2) determine an optimal raster resolution of grid simulation units at a fixed soil vector map scales, based on an assessment of the simulation units' data accuracy and redundancy metrics, and (3) construct relationship between soil vector map scale of polygon units and optimal raster resolution of grid units for DNDC modelling at regional scale. The results will serve as a reference for soil simulation unit conversion from polygon to grid format, in the support of soil carbon cycle modelling at regional scale.
The Tai Lake region
(118
First of all, paddy polygon unit datasets for DNDC simulation at
six soil vector map scales, e.g.
The paddy polygon unit datasets at the six map scales were developed by a Gis Linkage technique based on Soil Type (Yu et al., 2005, 2007a, b), namely PKB (Pedological Knowledge Based) method (Zhao et al., 2006), from soil vector maps at their corresponding map scales, respectively. The soil vector maps were compiled using a standard soil mapping system formulated as part of the Second National Soil Survey of China conducted in the 1980s (Office for the Second National Soil Survey of China, 1994). To the six map soils, soil species is the basic mapping unit for C5 and D2, soil family is for P5 and N1, while soil subgroup is for N4 and N14 (Yu et al., 2014). The soil properties attributed to all paddy polygons were derived from soil profiles, which were surveyed, compiled and authorized in the Second Soil Survey of China in 1980s (Shi et al., 2006). The number of representative soil profiles whose measured data were applied to attribute paddy polygons at C5, D2 and P5 scales totaled 1107, 136 and 127, respectively. The datasets were all taken from three books: Soils of County, Soils of District and Soils of Province, respectively. The paddy polygons at national map scale (N1, N4 and N14) were origined from 49 soil profiles described from the book “Soils of China” (Shi et al., 2006; Yu et al., 2014).
Secondly, paddy grid unit datasets for DNDC simulation were
developed from above paddy polygon unit datasets at the six map
scales. Each vector paddy polygon unit dataset was converted to
a series of paddy grid unit datasets of differing grid cell
sizes. The gird cell size ranged from a default size to a maximum,
with the size increment set to approximately 10 % of the
default. The default was determined by the soil vector map scale
and the lowest mapping unit size (
Finally, all simulation units rendered as vector (polygon unit) and raster (gird unit) datasets describing the soil properties, daily weather, cropping systems, and agricultural management practices of rice paddy fields, are required to initialize and run the DNDC model at regional scale (Yu et al., 2011, 2013). Each simulation unit has own data records specifically that were used as input for the DNDC modelling of SOC dynamics (L. Zhang et al. 2009; L. M. Zhang et al., 2009, 2012, 2014).
The DNDC (DeNitrification–DeComposition) model is a process-base model of carbon (C) and nitrogen (N) biogeochemistry in agroecosystems (Li et al., 1992a, b), it can simulate soil C and N biogeochemical cycles in paddy rice ecosystems, depending on a series of anaerobic processes being supplemented in the model (Li et al., 2002, 2004; Li, 2007).
For DNDC modelling of SOC dynamics, farming management scenarios
were compiled based on five assumptions from
L. Zhang et al. (2009) and L. M. Zhang et al. (2009, 2012, 2014), did not vary with the soil simulation unit
within counties. The DNDC modelling runs span the time period 1982
to 2000, duration of 19
To validate and assess performance of DNDC modelling, observed
values of SOC content acquired in 2000 from 1033 soil sampling
sites within paddy polygon units at C5 map scale, were used to
against modelling values (L. M. Zhang et al., 2014). The observed SOC
content of top layer (0–15
Simulated SOC density (SOCD,
Four indices of surface paddy soil, Paddy soil area (AREA, M ha),
number of paddy soil type (STN), the simulated SOC stocks (SOCS,
Tg) and average SOCD (ASOCD,
Variation of an index value (VIV, %) obtained from a grid unit
dataset (IV-
The optimal soil grid unit size for a polygon unit dataset conversion to grid unit dataset is the maximum grid cell size of which the two datasets are scaled identically. Statistical analyses were conducted by using the Excel and Origin 10 software.
Soil organic matter, clay content, pH and soil bulk density are all sensitive parameters as input for DNDC SOC simulations (Li et al., 2002; Levy et al., 2007; L. M. Zhang et al., 2012, 2014). The spatial distribution characteristics of these soil properties depicted by various simulation unit datasets differ from each other. The difference of the input parameter value affects uncertainty of the modelling (Valade et al., 2014; Zhu and Zhuang, 2014). A map scale or raster resolution decrease yielded a change in their estimated content (Tables 1–6), and a corresponding change in the simulated SOC (Table 7).
Weather data (precipitation, maximum and minimum air temperature) and farming management scenarios (sowing method, nitrogen fertilizer application rates, livestock, planting and harvest dates, etc.) variability among these simulation unit datasets for the purposes of this analysis can be neglected, because they were from the same weather and farming management county scale database (Yu et al., 2011, 2013) overlain with these soil polygon datasets. Change in soil type and their attributes as well as soil type area are the main source of SOC variability simulated by DNDC associated with the simulation unit scale and resolution (Yu et al., 2011, 2013).
The basic mapping unit's type, numbers of paddy soil type (STN) and polygon unit (SPN) as well as soil area (AREA) determined from the six paddy polygon unit datasets at different map scales, which describe the physical characteristics of these soil datasets, differ from each other (Table 7). For instance, four of the six paddy soil subgroups, Bleached, Percogenic, Degleyed and Submergenic paddy soil, do not get described in N14 polygon unit dataset but in other five datasets. The data scarcity should be one of the substantial causes of the uncertainties in modelling on regional scales (W. Zhang et al., 2014) did. And understandably, the C5 paddy polygon unit dataset containing the maximum numbers of soil polygon units, soil families and species (Table 7), is the most detailed and accurate database in the Tai Lake region (L. M. Zhang et al., 2009, 2012; Yu et al., 2011, 2013). That the IVs of STN, AREA, SOCS and ASOCD obtained in C5 dataset are considered to be the most believable in the region (Yu et al., 2011, 2013).
The IVs of SOCS and ASOCD for surface paddy soils simulated by DNDC
with the six polygon unit datasets display pronounced difference
from each other, as well (Table 7). In the main, the IV of SOCS
increased with decreasing of the map scale of polygon unit
dataset. The highest IV of SOCS was simulated with the N14 polygon
unit dataset, due to the largest area of the Hydromorphic paddy
soils with the highest SOCD (
The three paddy polygon unit datasets C5, D2, P5 are representative of regional scale digital maps, describing soil features at the county, district and province levels, respectively (Yu et al., 2013). The VIVs of the four assessment indices (STN, AREA, SOCS and ASCOD) determined from grid unit datasets and their parent polygon unit dataset, increases with increasing grid cell size (Fig. 3a–c). VIV magnitude and trend vary with grid cell size and by dataset and index. For instance, the VIV of STN from C5 or D2 datasets varies with grid cell size best described by an exponential curve (Yu et al., 2011), while the VIV from P5 varies as a logarithmic curve (Yu et al., 2014).
To the C5 polygon unit dataset and affiliated grid unit datasets,
VIVs of the four indices are all
Similarly, for D2 and P5 dataset conversion, only the VIV of ASOCD
is
The three paddy polygon unit datasets of N1, N4 and N14, describe soil features at the national scale (Yu et al., 2013). Generally, almost all VIVs of the four assessment indices from these grid unit datasets and their parent polygon unit datasets increase with increasing grid cell size except N14 (Fig. 3d–f).
For example, the VIVs of three index (SOCS, AREA and ASOCD) from
the N14 dataset conversion varies with grid cell size in the
diagram of random scatter except the STN index of soil subgroup
when the grid cell size ranges from 18 to 36
The results for N1 and N4 datasets conversion demonstrate that the
VIVs of ASOCD and STN are
Correlation analysis indicated a statistically significant
relationship between paddy polygon unit map scale (
At map scale of C5 (
By contrast, the grid units at the cell size of
The harmonized world soil database (HWSD), completed by
FAO/IIASA/ISRIC/ISSCAS/JRC in 2009, was produced at a cell size of
about
Considering Fig. 2 and Table 1 we see that the influence of the geomagnetic and magnetospheric terms is negligible. Furthermore, Eqs. (1) and (2) add no insight to the problem. We must therefore conclude that Phillips (1999) incorrectly supposed such a connection to exist.
In spite of this negative result, research will continue on this highly interesting question. For if it were to prove correct, then the consequences would be enormous to say the least.
Yu et al. (2014) did similar study by using similar method and
same basic data in same region as this study. A difference of
method adopted in Yu et al. (2014) from this study was that SOC
content (
The quadratic curve regression model (Eq. 8) revealed in this study
differ from a standard linear regression too, as Yu et al. (2014)
did, which describes the relationship between soil polygon unit map
scales and matched default grid cell sizes (Fig. 4). The quadratic
model implies that when the map scale is larger than
Almost all map scales of soil polygon unit datasets for China being frequently used are involved in this study, which were generated from the Second National Soil Survey of China. The six soil map scales were designed for soil mapping at different administrative levels including county, district, province and the whole country (Shi et al., 2006).
The Tai Lake region is a typical area in China where paddy soil prevails. Although it is located in the Yangtze Delta plain in East China, where rice fields are integrated with a high density of river or pond, garden and urban land, the spatial pattern of rice field distribution is similar to hilly or mountain regions where rice fields coexist with crop, grass, shrub and forest and urban land (Yu et al., 2011, 2013, 2014). We may assert with some degree of confidence that the knowledge obtained in this present study can be rolled out elsewhere in East and South China where distributes 95 % of rice filed in China (Li, 1992c).
While in the North and West China, soil vector mapping unit is larger in size than that of East and South China at various map scales, because of simpler natural conditions and reduced spatial variability. We may draw a conclude from it that the optimal grid cell size determined from the quadratic model (Eq. 8) can be smaller than the real optimal size in the region (Yu et al., 2011, 2013, 2014). The optimal grid cell size applying will result in a little redundancy of grid unit dataset, but not affect its accuracy matching to their soil polygon units' map scales. Although the quadratic model was obtained from a specific case study, and it would vary with the research region, the knowledge can be used as a guideline for soil unit conversion from polygon to grid, and for optimizing field sampling strategies, to support the regional simulation of SOC pool dynamics in China.
Within China a few administrative region extents are different
from those used here, which is caused by their history
anthropogeography and physical geography, resulting in additional
soil datasets with non-traditional map scales, such as
The DNDC model has been utilized to upscale estimates of SOC from the plot to region scale. For DNDC up-scaled utilization, a region is partitioned into many simulation units, e.g. soil vector polygon units or raster grid units, within which all properties are assumed to be as homogeneous as they are at plot scale. The homogeneity assumption is a possible major source of error when extending DNDC modelling from the plot to region scale. The homogeneity of simulation units is linked to soil polygon units map scale and grid units resolution, which has a strong influence on the results of SOC pool simulation.
Soil grid units are more often applied to SOC pool simulation, as
they are more easily manipulated for spatial model simulation,
geo-statistics and spatial analysis than soil polygon units. Most
of them are derived by data conversion from soil polygon units, but
the grid unit resolution choice varies by researcher even if they
are derived from a certain vector polygon unit dataset. An optimal
raster resolution matched with a certain map scale, for soil
polygon unit conversion to grid unit, was put forward in this
study. The optimal raster resolution is the maximum grid cell size
of which the soil grid unit dataset and the vector polygon unit
dataset are scaled identically. The optimal soil grid unit
resolution was found as
For the investigation and simulation of regional SOC pool, the quadratic curve model is more important to the soil polygon unit conversion at N4 less map scales than the other map scales. Although the quadratic curve model was revealed from a specific case study and would vary with the investigated region, the knowledge can be used as a guideline for soil assessment unit conversion from vector polygon to raster grid, optimizing field sampling strategies, and minimizing uncertainty of the investigation and simulation of regional SOC pool at different map scales further.
D. S. Yu and H. D. Zhang pondered the rationale of the method. X. Z. Shi collected the observed and simulated datasets. Y. L. Ni and L. M. Zhang performed the DNDC model simulation. H. D. Zhang and D. S. Yu prepared the manuscript with contributions from all coauthors.
We gratefully acknowledge support for the research from “Strategic Priority Research Program – Climate Change: Carbon Budget and Related Issues” (XDA05050507), the National Basic Research Program of China (2010CB950702), and the Natural Science Foundation of China (40921061).
Statistics of soil parameters input from
different resolution units at the map scale of
Statistics of soil parameters input from
different resolution units at the map scale of
Statistics of soil parameters input from
different resolution units at the map scale of
Statistics of soil parameters input from
different resolution units at the map scale of
Statistics of soil parameters input from
different resolution units at the map scale of
Statistics of Input Soil Parameters for
Different Resolution Unit at map scale of
Index values determined from DNDC simulations
with the paddy polygon units at different map scales in the Tai Lake Region
of China
*SOCS: SOC stocks of surface paddy soil; AREA: paddy soil area; ASOCD: average SOC density of surface paddy soil; STN: paddy soil type number; SPN: paddy soil unit number; S1: soil species; S2: soil family; S3: soil subgroup; S4: soil great group (Paddy soil).
The location of Tai Lake region.
Map of soil organic carbon
density (SOCD) simulated by DNDC from vector paddy soil units at different
map scales in the Tai Lake region of China.
(
VIVs varied with grid unit
resolutions at different soil unit map scales in the Tai Lake region of
China (VIV, Variation of an index
value; SOCS, soil organic carbon
stocks simulated by DNDC; AREA,
soil area; ASOCD, average soil
organic carbon density simulated by DNDC;
STN, soil type number;
Relationship between paddy polygon unit map scale and matched optimal grid unit resolution for the SOC simulation with DNDC in the Tai Lake region of China.